Final Delta

In the early 1970s, McDonnell Douglas set out to upgrade its Delta to carry satellites that were too heavy for the 2000 series, but not so large as to need an Atlas-Centaur. As introduced in 1975, the 3914 was its 2000 series counterpart with Castor IV (TX-526) strap-ons and a Star 37 third stage with a capacity to geosynchronous transfer orbit of 930 kilograms. In 1980 the third stage was replaced by the PAM-D. This Payload Assist Module was developed by McDonnell Douglas for use by Shuttle payloads in need of a perigee kick motor, but the Shuttle was late, and it had been decided to introduce the motor on the Delta - hence the 'D' qualifier. It incorporated a Star 48 motor. Perversely, as the PAM was considered to be part of the payload, the designation had a '0' to indicate the absence of a third stage. The 3910/PAM raised the capacity to geosynchronous transfer orbit to 1,090 kilograms. In 1982 a new second stage was introduced using the AJ-10-118K engine that had been developed for the Titan III's Transtage, which had two, as the AJ-10-138. The fact that this new stage had 30 per cent more propellant gave the 3920/PAM a capacity of 1,250 kilograms. The more powerful PAM-D2 (which used a Star 63 motor) was introduced later for heavier satellites.

Only three of the 38 Delta 3000 series dispatched between 13 December 1975 and 24 March 1989 had difficulties. On 13 September 1977 a 3914 carrying OTS 1 exploded at T+54 seconds, probably due to a burn-through of the casing of one of the strap-ons.3 A 3914 launched from Canaveral on 7 December 1979 suffered an apogee motor failure that stranded Satcom 3 in geosynchronous transfer orbit. On 3 May 1986 the first stage of a 3914 with GOES G shut down 71 seconds into the planned 223-second firing. This was the first premature shutdown for the RS-27 engine. The telemetry indicated that there had been two power surges immediately beforehand.4 The wreckage was salvaged, and the electronics that controlled the engine was examined for evidence of a short circuit. McDonnell Douglas built a high-fidelity simulator of the first stage to enable failure scenarios to be assessed, and it was concluded that the fault was a wiring harness.5A7 The investigation found that mechanical damage to wiring by vibration had - as Larry Ross, the chairman of the NASA Delta Review Board put it - prompted ''a rather significant electrical fault''. The battery current, ordinarily running about 9.5 to 9.8 amperes, at about 70 and some-odd fraction of seconds into flight, took a very significant excursion on the order of 188 amps.

It did that for about 6 milliseconds. It recovered to normal and stayed that way for about 900 milliseconds. It took another excursion of about the same magnitude, this time for about 13 milliseconds and once again recovered. The impact of that, and the direct cause of the failure, was a reduction in voltage available to relays that are in the aft end of the [stage]. Those relays hold open propellant valves, which feed propellants to the engines. Once sufficient voltage to keep those relay coils energised is lost, even momentarily, it is not possible to reactuate those valves.

At 71 seconds, the main engine shut down. Without attitude control, while still under the power of the three air-started strap-ons, the vehicle entered a flat spin, and by 77 seconds it had developed sufficient yaw to over-stress the shroud and shear it off, in the process demolishing the payload. The vehicle had completed a full 360-degree rotation before the range safety officer commanded its destruction.8 Several years earlier, the poly-vinyl chloride insulation on the wiring had been superseded by teflon insulation, and the shape of the wiring harness had allowed the vibration of launch to cause mechanical damage to the insulation. The harness was redesigned.9 On 26 February 1987 GOES 8 was dispatched, and despite a severe electrical problem with its apogee motor, was successfully inserted into geostationary orbit to give the National Oceanic and Atmospheric Administration a much needed second operational satellite.10


In 1978, President Jimmy Carter ordered that expendable launch vehicles be phased out in favour of the Space Shuttle which, starting in 1980, was to operate as the National Space Transportation System.11 The protracted development of the Shuttle prompted the offloading of some commercial satellites, and on 15 November 1980 a Delta 3910/PAM launched SBS 1, the first of the HS-376 satellites. The Shuttle Columbia made its debut on 12 April 1981, and after its fourth test flight in June 1982 was declared operational. When launched on 11 November 1982 for mission STS-5, it carried a pair of HS-376s mounted vertically in cradles towards the rear of its payload bay. A satellite launched on an expendable rocket would be spun up and released immediately, but the Shuttle was to deploy its satellites over a period of several days. As soon as the bay doors were opened, a clamshell shade was closed

The Delta carrying GOES G is destroyed.

The launch of STS-1.

Satellite Launch Cradles

The deployment of the HS-376 satellite Anik C2.

over each satellite to protect it from the Sun, and a heater was activated. The Shuttle faced its bay forward and tilted one wing downward, so that the satellites would be aligned with the velocity vector. Eight hours into the mission, SBS 3 was spun up by the turntable in its cradle and then ejected by a spring - and the National Space Transportation System was in the commercial satellite business. The satellite was on a 2.2-tonne PAM stage that could insert a 1.25-tonne satellite into geosynchronous transfer orbit, and after 45 minutes, while crossing the equator, the motor fired for 83 seconds and was released. At the apogee of the transfer orbit 6 hours later, SBS 3 fired its own thrusters to circularise. The next day, an Anik C satellite was similarly dispatched for Canada's Telesat.

Was this article helpful?

0 0

Post a comment