Atlas Centaur

The Centaur was a powerful upper stage that used a pair of RL-10 hydrogen-burning engines. On the first test of the Atlas-Centaur on 8 May 1962, the shroud collapsed early in the ascent.22 On 27 November 1963 the Centaur achieved orbit, but on 30 June 1964 it suffered a hydraulic failure. The flight on 11 December 1964 succeeded, but on 2 March 1965 the vehicle exploded on the pad. After two successes it made its operational debut on 30 May 1966 by dispatching Surveyor 1 to land on the Moon. On 10 August 1968, on its first attempt to insert a satellite into geosynchronous transfer orbit, the Centaur failed to reignite, stranding ATS 4 in parking orbit. On 30 November 1970 the shroud failed to separate, trapping OAO B. On 8 May 1971 a faulty gyroscope circuit in the autopilot prevented the Centaur gimballing its engines, with the result that Mariner 8 was dumped in the Atlantic.23 Intelsat 406 was lost on 20 February 1975 when the Atlas suffered an electrical failure during separation, and Intelsat 4A5 was lost on 30 September 1977 when one of the turbopumps in the Atlas failed at T+55 seconds.24

When the Atlas-G with its stretched core was introduced on 9 June 1984, the Centaur exploded as a result of an oxygen leak caused by an anomalously violent separation, and Intelsat 509 was lost.25 The first FLTSATCOM satellite for the Navy was put into geosynchronous transfer orbit by an Atlas-Centaur in 1978, but when the fifth satellite in this series was launched on 6 August 1981 the fibreglass shroud collapsed.26 Although the Centaur put the payload into geosynchronous transfer orbit and the Star 37F apogee motor circularised the orbit, the satellite had been badly damaged by the shroud and had to be written off.27 With the initial FLTSATCOMs surviving beyond their five-year design life, the first of the Block-II series, FLTSATCOM 6, lifted off on 26 March 1987 on the penultimate Atlas-G into driving rain from a 2,500-foot overcast. When it strayed off course at T+51 seconds the range safety officer destroyed it.28 John W. Gibb, the manager of the Atlas-Centaur project office, reported that this launch had not violated the long-established rules that allowed a launch in rain, prohibiting doing so only if there was a thunderstorm within 5 miles of the pad or anvil clouds within 3 miles of the pad.29 In this case, the countdown had been 'held' to await acceptable conditions. As the debris had fallen close offshore it was readily recovered and the investigation, led by John R. Busse, soon found a piece of the aerodynamic shroud with a number of pin-hole punctures burned through it, indicating that the vehicle had been hit several times by lightning.30 This was confirmed when the digital computer unit was located and showed that the last command gimballed the engines hard over.31 This was not a software error, the false command was the result of ''a single random upset'' in the computer's memory induced by the intense electric field of the lightning strike. The erroneous command had been issued at T + 38.3 seconds, and the aerodynamic loads had caused the vehicle to break up at T+50.7 seconds. The destruction command had finished the job. It was concluded that the launch had been a ''vehicle no-trial'', because the Atlas had been dispatched into an environment in which it was not designed to survive.32,3V4 Busse dismissed NASA's claim that the launch had not violated the commit criteria. He said the basic failure was a ''missed call'', and criticised the lack of coordination between the weather forecasters and the launch team.35 NASA upgraded its local weather monitoring (particularly of the electric fields) to enable it to make a better assessment of when an acceptable 'window' might occur in poor weather. The loss of the Atlas was in marked contrast to Apollo 12, which was launched on 14 November 1969 into an overcast of stratocumulus in defiance of a rule that no vehicle be launched into a thunderstorm. No sooner had the 360-foot-long Saturn V cleared the 400-foot tower than it vanished into the murk. At T + 36.5 seconds, at an altitude of 6,100 feet, the vehicle was hit by a bolt of lightning that discharged to ground down the exhaust plume. A reverse electrical surge damaged some of the instrumentation in the spacecraft and caused its fuel cell system to disconnect from the power buses, and 16 seconds later a cloud-to-cloud bolt disrupted its inertial guidance. Fortunately, the Saturn V had its own

Atlas-Centaur, 26 May 1977 with Intelsat 4A5.

The launch of Apollo 12 into a thunderstorm, and the resultant lightning strike half a minute later.

power supply and guidance, and flew on unscathed.36 As this was the first time that Walter Kapryan (the launch director of the Kennedy Space Center) and Gerry Griffin (the flight director in Houston) had supervised an Apollo launch, they promptly became known as "the rookies who launched into a thunderstorm".37


When the Atlas became a priority, the Air Force ordered the parallel development of a completely different design, in case the Atlas proved impractical. The contract went to the Martin Company. By this time, the engineering database had advanced sufficiently to make air-starting a liquid rocket an acceptable challenge, and it was decided to employ an efficient tandem configuration. The first stage had a twin-chamber Aerojet LR-87 engine and the second had a single-chamber LR-91, both burning kerosene and oxygen. The engines were to gimbal to steer the vehicle in powered flight, and verniers were to refine the trajectory of the second stage after its main engine shut down. Development was rapid, and in February 1959 the first stage was tested using a mock-up second stage. Ten full-range flights followed in 1960. Like the Atlas-F, the Titan was to be stored vertically in a silo and raised for firing. Being more advanced, it could have been expected to complement and later supersede the Atlas, but in 1963 the Air Force decided to withdraw both. Unlike the Atlas, the Titan was junked. The Titan II, on which work started in 1960, was a very different story. Although it benefitted from lessons learned with the Titan I, it was a wholly new design and a variety of technological developments simplified its construction. In this case the engines burned hydrazine and nitrogen tetroxide which, being hypergolic, burned on coming into contact, thereby eliminating the igniter. The fact that these propellants were not cryogenic enabled the missile to be stored ready to fire from its silo literally at one-minute's notice. Development was rapid, and after a test programme that began with the debut launch in March 1962, the Air Force began the deployment of three 'wings', each of 18 missiles, in 1963.

In addition, the Air Force intended to use the Titan II as a space launcher, and

Hypergolic Propellant
Titan II and early III models (left to right): Titan II ICBM; Titan II, 8 April 1964 with Gemini 1; Titan IIIA, 11 February 1965 with the Lincoln experimental communications satellite; Titan IIIB, 6 November 1968 with the 17th KH-8 satellite; Titan 34B, 24 April 1981 with lumpseat 6.

NASA ordered a dozen adapted for use in the Gemini programme in 1964-1966. The Titan IIIA was a 'stretched' Titan II fitted with a third stage. The first test from Pad 20 at Canaveral on 1 September 1964 went well until a pressurisation failure caused the premature shut down of the third stage, known as the Transtage. In 1966, after three successful flights of the Titan IIIA, the Air Force switched over to the Agena-D for the simple reason that the mass-produced stage was more cost-effective for the envisaged missions. The first of the Titan IIIB vehicles was successfully dispatched from SLC-4W at Vandenberg on 29 July 1966. All but two of a total of 54 launches over the next 20 years successfully inserted KH-8 'close look' reconnaissance satellites into low polar orbit.38 The losses were on 26 April 1967 when the second stage lost thrust, and 26 June 1973 when the Agena failed.

Meanwhile, the Titan IIIC was introduced by augmenting the Titan IIIA with a pair of five-segment solid rocket boosters. The inaugural flight from Pad 40 at Canaveral was a success with a dummy payload. On the second flight on 15 October 1965, however, the Transtage broke up. Nevertheless, all but three of 35 fired from Canaveral between 1966 and 1982 successfully deployed a variety of payloads. The failures were 26 August 1966 when the shroud failed at T + 78 seconds, 20 May 1975 when the Transtage lost attitude control following a power failure to its gyroscopic platform, and 25 March 1978 when a hydraulics pump in the second stage failed. The Titan IIID was essentially the Titan IIIB with five-segment boosters. The first lifted off from SLC-4E at Vandenberg on 15 June 1971, and 22 successfully deployed KH-9 and KH-11 reconnaissance satellites by 1983. The Titan 34B was the Titan IIIB with an enlarged shroud that enclosed both the Agena and either an SDS data-relay or a Jumpseat electronic intelligence-gathering satellite. Of 14 launched from SLC-4W between 1971 and 1987, all but one successfully inserted its payload into an orbit with a high apogee above the northern hemisphere; the failure was on 16 February 1972 (the second flight). Meanwhile, NASA had introduced the Titan IIIE, which

Titan Iiib

Titan IIIC and Titan IIID models (left to right): Titan IIIC, 18 January 1967 with a dispenser of eight satellites for the Initial Defense Communications Satellite Program; Titan IIIC, 14 December 1975 with the 5th IMEWS (the last of the first model of DSP); Titan IIID, 15 June 1971 with the 1st KH-9 'Big Bird'; Titan IIID, 14 June 1978 with the 2nd KH-11.

Titan IIIC and Titan IIID models (left to right): Titan IIIC, 18 January 1967 with a dispenser of eight satellites for the Initial Defense Communications Satellite Program; Titan IIIC, 14 December 1975 with the 5th IMEWS (the last of the first model of DSP); Titan IIID, 15 June 1971 with the 1st KH-9 'Big Bird'; Titan IIID, 14 June 1978 with the 2nd KH-11.

was a Titan IIIC with a Centaur upper stage. Although the Centaur failed on the vehicle qualification flight on 11 February 1974, over the next three years six were successfully launched from Pad 41 with a pair of Helios, a pair of Viking, and a pair of Voyager spacecraft.

The Titan 34D was the Titan 34B augmented with longer strap-ons with five-and-a-half segments per booster and either a Transtage or a two-stage IUS instead of the Agena. Its introduction on 30 October 1982 from Pad 40 at Canaveral marked the debut of the IUS, which successfully delivered a pair of DSCS communications satellites to geostationary orbit for the Department of Defense. The first Titan 34D to be launched from SLC-4E at Vandenberg was on 20 June 1983. Of 15 launched with a variety of payloads by 1989, all but three were successful. On 28 August 1985 a Titan 34D tumbled and the range safety officer destroyed it.39,4° Not only was this the first failure for this variant, it was the first such mishap in 18 years of Titan operations at Vandenberg. The investigation concluded that the evidence indicated an oxidiser leak and the consequent failure of a turbopump subassembly when its pinion gear broke as a result of the loss of gear cooling or lubrication. This caused the premature shut down of one of the two engines on the core stage, and because the vehicle required both engines to maintain controlled flight, it tumbled.41 This incident highlights an important consideration in launch vehicle design - a vehicle with more engines will lose a smaller fraction of thrust if one engine fails, thus (as for aircraft) launchers with more engines are perceived as being more reliable. However, if the vehicle's thrust margin is so low that it requires all of its engines to operate, then the probability of a launch failure will be greater for a many-engined vehicle than for one that has only one engine. By the end of 1985, however, the Air Force was phasing out the Titan, and had only a few left.

A Titan IIIE-Centaur lifts off on 20 August 1976 with Voyager 2.


On 29 April 1959 (six months after its establishment) NASA ordered 12 Thor-Able launch vehicles, which it renamed the Thor-Delta, and it also asked Douglas to improve the inertial system to provide more accurate guidance during the ballistic coast leading up to the release of the third stage.42 The maiden launch on 13 May 1960 from Pad 17 at Canaveral failed because the second stage suffered a loss of attitude control, but the next mission on 12 August 1960 successfully placed the Echo passive relay 'balloon' satellite into orbit. The last of the batch was fired on 18 September 1962 and, apart from the first, they all succeeded with a variety of pay-loads which included the first TIROS weather satellite and Telstar, the first active-repeater communications satellite.43 By this time, of course, a delighted NASA had ordered another batch of rockets. Fortunately, the potential of the Thor enabled Douglas to exploit the 'building block' approach to undertake a programme of progressive upgrades which were designated alphabetically. Two launches of the Delta-A in October 1962 tested the more powerful MB-3-II first-stage and the AJ-10-118 second-stage motor, which had a restart capability, and on 13 December 1962 the Delta-B introduced the AJ-10-118A with a longer burn time and better guidance. Nine vehicles of this type were launched with a payloads that included Syncom, the first communications satellite to be placed into geostationary orbit.44 The only failure was on the final launch, on 19 March 1964, when the third stage generated insufficient thrust and stranded Explorer 20 in the wrong orbit.

In 1963, the Delta-C had the MB-3-III engine and the AJ-10-118D, and introduced the X-258 Altair II third stage. Of 12 fired between 27 November 1963 and 22 January 1969 only one failed. On 25 August 1965 the early ignition of the third stage in the post-boost coast doomed OSO C. Meanwhile, the Air Force had augmented its Thor-Agena by strapping three Castor I (TX-33) solid rockets to the first stage, and this strategy was copied for the Delta-D, which was introduced as the Thrust Augmented Thor (TAT) Delta on 19 August 1964. On 6 April 1965 one put the 'Early Bird' communications satellite in geostationary orbit above the Atlantic.

Titan 34D models (left to right): Titan 34D, 30 October 1982, with an IUS carrying the 15th and final DSCS II and the 1st DSCS III; Titan 34D, 10 May 1989 with a Transtage carrying the 6th and final Chalet satellite; Titan 34D, 18 April 1986 with the 20th and final KH-9.
Delta Early Bird Satellite

A Thor-Delta launches Telstar 1 on 10 July 1962. The inset shows a scene from the first transatlantic TV relay.

A TAT Delta launches the Early Bird on 6 April 1965.

Next, the Thor was stretched for greater endurance and augmented by three of the more powerful Castor II (TX-354) strap-on solid rockets, and the second stage was increased in diameter to carry propellant to triple its duration and fitted with the more powerful AJ-10-118E. Termed the Delta-E, six of these Improved TAT Deltas were launched between 6 November 1965 and 20 April 1967, with no failures. A second form utilised the United Technologies FW-4D solid motor as its third stage. Only one of 18 flights between 1 July 1966 and 1 April 1971 had a problem. This was on 26 August 1966, and it was not really a launch vehicle issue as the Intelsat 2F1 satellite stranded itself in geosynchronous transfer orbit when its built-in apogee motor shut down after just 4 seconds.45 The Delta-F was the FW-4D form of the Delta-E without strap-ons, but there proved to be no requirement for it. The Delta-G was a two-stage version of the Delta-E using the older Castor I strap-ons. It was used in 1966 and 1967 for two recoverable Biosatellites. The Delta-H was the Delta-G without strap-ons, but again there was no need for it. The Delta-J, which was the Delta-E with a Star 37 third stage, was used only once, in 1968. The Delta-K never got off the drawing board.

The next major advance was the Long Tank TAT Delta, in which the first stage was made cylindrical rather than conical, and stretched by 15 feet. The Delta-L, -M and -N differed only in their third stages: the Delta-L had the same third stage as the Delta-E; the Delta-M had the same third stage as the Delta-J, and the Delta-M-6 was the Delta-M with three extra Castor II strap-ons (for a total of six, thereby giving rise to the name Super Six); the Delta-N was a two-stage version of either the Delta-L or the Delta-M, and the Delta-N-6 was a two-stage Delta-M-6. On the maiden launch of the Delta-L on 27 August 1969 a hydraulic oil relief valve leaked due to vibration, resulting in loss of hydraulic pres-

Delta 1913, 10 June 1973 with Explorer 49.

sure which permitted the engine nozzle to develop uncontrolled gimballing and attitude excursions, and at T + 383 seconds the range safety officer destroyed the vehicle and its payload, which was the Pioneer E satellite that was to have been placed into solar orbit to report on the solar wind.46 When the first Delta-M was launched on 19 September 1968 a pitch-rate system malfunction was noticed about 20 seconds after lift-off, the first stage began to break up at T+102 seconds, and 6 seconds later the range safety officer destroyed it and its Intelsat 3F1 payload.47 In the case of the Delta-M on 26 July 1969, the third stage motor's casing either ruptured or there was nozzle failure, with the result that Intelsat 3F5 was stranded in the wrong orbit.48 When the first Delta-N lifted off from SLC-2 on 21 October 1971 with the ITOS B weather satellite, the second stage suffered an oxygen leak.49 The pitch and yaw jets pulsed to counteract the force of this venting, maintaining the proper attitude until their gas was expended, whereupon the vehicle tumbled.50 On 16 July 1973 a two-stage vehicle with ITOS E suffered an attitude control malfunction 270

Delta 1913, 10 June 1973 with Explorer 49.

seconds into the second-stage burn following an abrupt fall in output from the hydraulic pump caused a loss of hydraulic pressure that disabled the thrust vector control system.51

Meanwhile, in April 1967 Douglas had merged with McDonnell, to form McDonnell Douglas. Having advanced far through the alphabet to accommodate the recent clutch of very similar variants, it was decided to switch to a numerical system that would provide a more structured scheme. The four digits specified in turn, the first stage, the number of strap-ons, the second stage and the third stage. In the 1000 series (which ran to the

Delta 2313, 19 January 1974 with Skynet 2A.

1410, 1604, 1900, 1910, 1913, 1914 variants) there could be 4, 6 or 9 Castor II strap-ons, the second stage engine was either the AJ-10-118F or the TR-201 (the latter being basically the descent engine of the Apollo Lunar Module with its throttle fixed at 100 per cent), and the third stage (if present at all) was one of several Star 37 solid motors. It introduced the Extended Long Tank configuration that was dubbed the Straight Eight because the upper part of the vehicle continued the 8-foot diameter of the first stage. This time the two-stage form was of use because the increased capability of the second stage enabled it to deliver a heavier satellite than the previous three-stage vehicle.5V3 Eight were launched between 23 September 1972 and 21 June 1975 with a variety of payloads.

The 2000 series marked a major upgrade that introduced a new power plant for the first stage. The Rocketdyne RS-27 combined the control system of the MB-3 with the turbopump, turbine, gas generator, valves and thrust chamber of the H-1 used in the first stage of the Saturn I series.54 Of 44 launches between 19 January 1974 and 6 October 1981 there were two failures. On the maiden flight (which was a 2313) a fragment of conductive contaminant was shaken loose by vibrations, causing a short circuit in the electronics of the second stage and, as a result, Skynet 2A was released into a low orbit that decayed after several days. The investigation concluded that the insulation coating on the printed circuit boards had been substandard and the contaminant had welded across the exposed ends of component leads.55 On 20 April 1977 the third stage of a 2914 was released before the rotator on the second stage had spun it up for stability, causing it to tumble.56 Nevertheless, by any standard, over the years the Delta had been a reliable launcher for small-to-medium payloads.

+1 0

Post a comment