September-December 1980:launch vehicle failures

Since the launch of Sputnik 1 on 4 October 1957, the spacefaring nations, especially the United States and the Soviet Union, have spent untold billions of dollars and roubles developing and producing the multitude of expendable launch vehicle models which have taken instruments and men into space. In the United States in particular, the number of models of expendable launch vehicles has added to the expense of space exploration and has probably reduced the overall reliability of the boosters. In 1973 alone, 25 missions were flown using 10 booster variations.

Methods, Definitions and Clarifications

Prior to discussing specific vehicles and their performance, it is important to define exactly what the term “vehicle failure” means in terms of this article. For this purpose, launch vehicle failure includes all launches which were intended, but failed, to place all or part of the vehicle into an orbit around the Earth, or launch that vehicle into more distant space, in such a way that the object or objects launched would have been awarded an international designation if the launch had been successful. Thus, only launches which failed to qualify for an international designation, but would have so qualified if successful, are included as failures.

Having stated the rule, we must now discuss the inevitable exceptions. Fortunately, they are few enough in number so they can be briefly discussed here. The Atlas Able 4 vehicle lost during a static test on 24 September 1965 is included because it resulted in the complete destruction of an already scheduled launch attempt. A failure which occurred in such a way that the launch could be scheduled later with the same booster would not be included as a failure. An example is the explosion on 21 January 1959 of the Agena stage of a Thor-Agena A. The Thor was saved and thus the mission is not counted as a success or failure but is listed for informational purposes.

Building Blocks of Success

The early years of space flight were marked by large numbers of launch vehicle failures. These failures were, however, the foundation upon which later successes were built. In its lifetime, the Vanguard programme achieved 3 successful satellite launches in 11 attempts for a success record of 27%. The experience gained in the development of the Vanguard was used to create the second stage of the Thor Able booster which achieved 4 successes in 7 flights for a 57% success rate. That programme in turn was a stepping-stone to the remarkably successful Delta launch vehicle which has had 133 successful launches in 140 flights achieving 95% success.

Similar progress was made with the Thor Agena. That programme grew from the Thor Agena A vehicle which scored 9 successful launches in 15 attempts (60%) to the Long Tank Thrust-Augmented Thor Agena D which launched 39 payloads into orbit in 41 attempts for a 95% success rate.

In the five years following the first satellite launch the United States made 167 launch attempts and succeeded 114 times for a success rate of 68%. In the five year period including the years 1968 through 1972, 187 launch attempts were made, with 177 succeeding, for an average of 95%. Prior to 1965 the United States had never achieved a yearly success average of 90% or above. After 1965 the United States never dropped below a 90% yearly success rate.

Commencing in the mid-1960’s,several booster programmes were initiated by the United States which maintained a 100% success rate. These vehicles included the Saturn I and IB (13 launches), the Saturn V (13 launches), the Titan II-Gemini (12 launches), the Thor Burner 2 (21 launches) and the Titan IIID (16 launches). The Thor Burner 2 and Titan IIID programme continue to this day. Without the failures of the early years, these remarkable records would not have been possible. The fact that no American astronaut has been lost during a launch, or even had a mission aborted, is also a direct result of the many lessons learned earlier.

Launch Vehicle Genesis

The world became painfully aware of the uncertainties involved in rocket development on 6 December 1957. Although the news media had carried numerous stories and photos of military missiles exploding on or above the launch pad prior to that date, the ill-fated launch pad explosion of Vanguard TV-3 on that day had an enormous impact on the future US space programme. The fact that TV-3 was a development flight seemed to make little difference to the public and Congress.

Between 6 December 1957 and 22 June 1959, the Vanguard vehicle suffered 8 failures and 3 successes. On 5 February 1958, TV-3BU suffered a control system malfunction after about 57 seconds causing the vehicle to break up. The next six failures involved problems at the point of initial separation or after, indicating that the 1st stage defects had been corrected but that much work remained to be done to resolve the continuing upper stage anomalies.

The proficient design team under Wernher von Braun reaped the well deserved glory of launching America’s first satellite (a feat which they were probably equipped to do a year or two earlier) but fared only somewhat better in their overall early success rate. The first, third and fourth launchings of the Juno I (Jupiter C) were successful. The second, fifth and sixth, however, all encountered upper stage problems indicating that the technology required for smooth operation of a multi-stage vehicle had yet to be perfected. The Juno II was the Army Ballistic Missile Agency follow-up launch vehicle to the Juno I. Between December 1958 and November 1960, this Jupiter-based satellite booster fired successfully in 5 out of 10 tries for a 50% average. On 16 July 1959, the attempted launch of Explorer S-1 failed when the Juno II guidance system failed after 5 seconds and the rocket proceeded to turn 180 degrees just above the launch pad and smash itself into the ground. The other four failures all involved the old problem of upper stage malfunction. The lessons learned in these launches paid dividends when the von Braun team moved on to the development of the Saturn family of boosters.

The Thor Able was the first US launch vehicle to achieve a percentage of success better than half. Out of seven launches,four were successful. This statistic is somewhat misleading,however, since one of these successes was the Pioneer 1 shot of 11 October 1958, when that probe reached 70,000 miles into space but returned to Earth after not quite achieving the velocity required to allow it to escape the Earth’s gravity. The three Thor Able launch failures included the Pioneer launch of 17 August 1958 in which the first stage exploded after 77 seconds and the Pioneer 2 Moon probe of 8 November 1958 where the third stage failed to fire.

The last Thor Able failure involved the attempted launch of Transit 1A in which the third stage also failed to fire.

The last of the very early launch vehicles was the Atlas Able. This booster was the only major category of satellite launcher that failed in every attempt. Atlas Able was utilized in a series of four attempts to fire probes at the Moon. Two Atlas Able 4 rockets were built and the first exploded in a static test on 24 September 1959, ending any chance of launching Pioneer P-1. The second Atlas Able 4 launched Pioneer P-3 on 26 November 1959 and was lost when the payload shroud fell off and the second stage failed to fire. The Atlas Able 5A launch with Pioneer P-30 was aborted when the second stage failed to fire and Atlas Able 5B with Pioneer P-31 exploded after 70 seconds of flight.

The Second Generation

The year 1960 saw the introduction of two launch vehicles that became workhorses of the first two decades of the space age. The first launchings of the Delta and the Scout were made that year. The Delta was a direct descendant of the Vanguard and the Thor Able. As such, it had the advantage of all the development work that had previously gone into those programmes. The launch of Delta 1 failed on 13 May 1960 when a 2nd stage attitude control failure occurred. Following that flight, however. Delta performed successfully for 22 missions until Delta 24 suffered a premature 3rd stage cut-off in March 1964. From that date to the end of 1978, the Delta programme suffered only five further complete launch failures in 123 attempts. A marginal booster failure occurred with the launch of Delta 86 on 21 October 1973 when the 2nd stage tumbled and ITOS E failed to orbit but the upper stage booster apparently did achieve a brief orbit. Of the complete failures, three occurred because of problems in the main booster and the other four suffered upper stage malfunctions. This record certainly indicates that the early problems with multistage boosters had been solved. The success rate in eighteen years of Delta booster history stands at 95%.

As of the end of 1978, the Scout launch vehicle had performed successfully 85% of the time, launching satellites into orbit in 68 of 80 attempts. The Scout vehicle has been flown by both NASA and the USAF. As a USAF launch vehicle it was called the Blue Scout. The NASA launches are well documented, but the USAF launches, especially during the years 1962 and 1963 are still shrouded in some mystery. The first three Scout failures were NASA attempts in 1960 and 1961. Scout 3 failed to launch Explorer S-56 on 4 December 1960 when the 2nd stage failed to ignite. Scout 5 failed on 30 June 1961 when the 3rd stage did not fire and Explorer S-55 failed to reach orbit. The guidance system failed on Scout 8 carrying Mercury-Scout 1 tracking equipment and the booster was destroyed after 30 seconds.

These NASA failures were followed by a series of seven USAF failures. The exact reason for 4 of the 7 failures appears at this time to be undisclosed. Fourth stage malfunctions probably caused the loss of Transit 5A2 on 6 April 1963 and OV3-5 on 31 January 1967. The second stage exploded in an attempt to launch ESRS on 26 June 1964. Undisclosed malfunctions caused the loss of USAF meteorological payloads on 24 May 1962, 26 April 1963, 27 September 1963 and Solrad 4B on 26 April 1962. On 29 May 1967 ESRO 11A was lost when the 3rd stage of a Scout failed. After that launch failure, there were 31 successful Scout launches before the failure of Explorer DAD A and B when there was a third stage nozzle failure on 5 December 1975.

Also introduced in 1960 was the Thor AbleStar. That launch vehicle flew successfully in 14 of 19 flights between 1960 and 1965 for a 74 percent success rate. The third Thor AbleStar failed to orbit Courier 1A when the first stage malfunctioned and the booster had to be destroyed after 2 minutes on 18 August 1960. Transit 3A/Greb 2 suffered a second stage failure causing booster destruction by the Range Safety Officer on 30 November 1960. Composite 1 (carrying five payloads) and Anna 1A both suffered second stage failures and were lost on 24 January 1962 and 10 May 1962 respectively. The final failure resulted in the loss of Transit 5BN3 on 21 April 1964 and was caused by human error attributed to the flight control team.

The Thor Agena

The Thor Agena launch vehicle combinations will take a prominent place in the history of expendable boosters. Between 1959 and the end of 1972, 186 launchings were made with the various Thor Agena configurations. Of those launch attempts, only 22 failed to orbit their payload for an overall success ratio of 88%. This figure is deceiving, however, since 12 of those failures occurred in the Discoverer programme utilizing the early Thor Agena A and Thor Agena B vehicles. Some authorities attribute a thirteenth failure to the Discoverer programme and indicate that they believe Discoverer 1, launched on 28 February 1959, never reached orbit although it received the international designation 1959 Beta.

What is particularly noteworthy is the ever increasing percentage of successes achieved by the various Thor Agena combinations. The Thor Agena A was launched 15 times suffering 6 failures, yielding a success rate of 60%. The Thor Agena B failed seven times in 44 attempts, achieving an 84% success rate. The Thor Agena D, first launched in 1962, failed in 3 out of 21 attempts for 86% success. The Thrust-Augmented Thor Agena D, launched first in 1963, succeeded in 59 of 63 attempts for a success rate of 94%. The final version, the Long Tank Thrust-Augmented Thor Agena D, failed only twice in 43 attempts, being successful 95% of the time.

The Thor Agena A programme suffered four failures related to the Agena second stage development. The fifth failure was caused by a malfunction in the first stage guidance system and resulted in the destruction of Discoverer 10 after 52 seconds of flight.

The launch attempts of the Thor Agena B resulted in seven failures of which six can be attributed to malfunctions in the Agena stage. The seventh failure occurred when the Range Safety officer was forced to destroy Discoverer 27 after one minute of flight. The three failures in the Thor Agena D launches occurred on 26 April 1963, 9 November 1963 and 2 September 1965. The reasons for the first two failures are undisclosed and the latter flight suffered an early guidance failure. It does appear, however, that the April attempt probably failed at the point of insertion, possibly because of insufficient thrust.

The four unsuccessful attempts of the Thrust-Augmented Thor Agena D appear to be the random failures which seem to be inevitable even in the most reliable of the expendable boosters. Undisclosed causes produced the failures of 24 March 1964 and 4 May 1966. Guidance failure caused the RSO to destroy the vehicle launched on 28 February 1963. The failure of 18 March 1963 was caused by a malfunction in the second stage.

The two failures of the Long Tank Thrust-Augmented Thor Agena D were caused by problems early in their flights. The first launch resulted in the loss of Nimbus B/Secor 10 on 18 May 1968 after 2 minutes. An unidentified payload was lost on 17 February 1971 after forty seconds in the air.

The Atlas Agena and other Atlas launches

The Agena second stage combined with the Atlas booster was introduced in 1960. As with the Thor Agena combination, the Atlas was used with all three Agena versions. In all, up to the end of 1978, 109 Atlas Agena vehicles were launched. There were twelve failures among these attempts for a success rate overall of 90 per cent. As with the Thor booster, much greater success was achieved with the later vehicle combinations.

The Atlas Agena A was launched four times with two failures for a 50% success rate. On 26 February 1960, the first Atlas Agena A failed when the second stage did not separate after a retro-rocket explosion. On 11 October 1960 an umbilical disconnect problem caused a later second stage failure. These failures resulted in the loss of Midas 1 and Samos 1 respectively.

The Atlas Agena B was launched 29 times with five failures for a success rate of 83%. On 9 September 1961, Samos 3 was lost when the launch vehicle exploded on the pad. Another Samos was lost on 22 November 1961 when the second stage of the booster failed. A NASA launch failed on 22 July 1962 when the Mariner 1 flight to Mars ended after five minutes following a guidance failure resulting in the RSO destroying the booster. Midas ERS 3 & 4 were lost on 17 December 1962 when the Atlas Agena B failed for an undisclosed reason and Midas ERS 7 & 8 failed on 12 June 1963 when the vehicle exploded just after launch. Partial failures occurred when Rangers 1 and 2 were left stranded in Earth orbit.

By far,the most successful of the Atlas Agena boosters utilized the Agena D second stage. Between 1963 and the end of 1978, 76 of these vehicles were launched and 71 orbited payloads for a success rate of 93%. Ironically, most of the 5 failures in this booster can be attributed to the Atlas first stage. Close-look satellites were lost on 8 October 1964 and 12 July 1965 when the RSO destroyed the booster in flight and the BMEWS E early warning satellite was lost on 4 December 1971 when a first stage guidance failure lead to the destruction of the vehicle by the RSO. The other two failures occurred in NASA launches when the target vehicles for GT-6 and GT-9 were lost. GTA-6 failed on 25 October 1965 when the Agena stage exploded after 6 minutes of flight and GTA-9 was lost on 17 May 1966 when the Atlas control system failed.

Malfunctions in Atlas boosted launch vehicles have occurred on other occasions as well. On 25 April 1961,NASA lost MA-3 when Atlas 100D suffered a control failure at lift-off and was destroyed by the RSO at 30 seconds after launch. Atlas 172D failed to orbit OV1-1 when separation failed to occur and pod injection failed on 21 January 1965. OV1-3 was lost on 27 May 1965 when the Atlas D booster exploded after 2 minutes. On 12 April 1965, an Atlas F failed to launch an unidentified payload when the Atlas vehicle malfunctioned. The only other identified Atlas failure (excluding the Centaur vehicles considered later) occurred when the Atlas was mated with the Burner second stage. On 16 August 1968, the Atlas Burner second stage failed to orbit 12 small geodetic satellites when the protective shroud surrounding the Burner 2 stage failed to separate precluding the second stage from operating. Between 1958 and the end of 1978, 30 launches were made using the Atlas B, D & F vehicles with four complete failures giving a success rate of 87%.

Centaur

The Centaur second stage was first flown with Atlas 104D in a suborbital attempt on 8 May 1962. That flight ended in failure when the nose fairing failed and the vehicle exploded after 55 seconds. Following that launch attempt, 48 Atlas Centaur launches succeeded in 88% of the flights attempted, resulting in only six failures. Two R&D flights failed when AC-3 had a premature second stage shutdown on 30 June 1964 and AC-5 exploded on the launch pad on 2 March 1965. Four operational failures occurred later. On 30 November 1970, OAO-B was lost when AC-21 suffered a problem in the nose fairing which caused the spacecraft to fail to separate. Mariner H was lost on 8 May 1971 when the gyros in the Centaur stage malfunctioned and AC-24 failed. AC-33 failed at separation on 20 February 1975 resulting in the loss of Intelsat 4(F-6). On 29 September 1977, Intelsat 4(F-4) was lost when the booster failed after 55 seconds on AC-43.

On seven occasions the Centaur second stage has been mated with the Titan IIIE booster. All of these firings were successful except for the first R&D shot on 11 February 1974 carrying a Viking test spacecraft and a Sphinx satellite. That vehicle failed when the Centaur stage failed to ignite. The Titan 3E-Centaur launches failed in one attempt of seven for an 86% success rate.

Titan as a booster

The Titan has been used in several other variations since the mid-1960’s and has achieved a consistently high success rate. During 1964 and 1965, the Titan IIIA was tested four times with the only failure occurring on 1 September 1964 when the transtage cut-off prematurely.

The Titan IlIC was first launched in 1965 and has since been fired 30 times. The two failures have resulted in a success rate of 93%. The first failure ended in the loss of DCSP B-1 on 26 August 1966 when a failure in the attitude control system resulted in the destruction of the booster after 82 seconds. The second loss occurred on 25 March 1978 when a second stage failure caused the vehicle and DSCS 2H and 2I to be destroyed eight minutes after launch.

The Agena D was first mated with the Titan in 1966. The Titan 3B Agena D has since been fired 57 times with only four failures. The booster has been successful 93% of the time. Failures occurred on 26 April 1967, 16 February 1972,20 May 1972 and 28 June 1973. The cause of the 1967 failure is undisclosed as of 1980, but the other three failures all apparently were related to malfunctions in the Agena stage.

Planetary Failures

Several partial failures have occurred which are considered successes for the purpose of this article. These five launches were all attempts to escape Earth’s gravity. They resulted,however, in Earth orbits only. A Thor Able 1 launched Pioneer 1 on 11 October 1958 and that Moon probe reached a height of 70,000 miles but had insufficient thrust to reach the vicinity of the Moon. Pioneer 3 was launched by a Juno II on 6 December 1958 and also had too little thrust to reach the Moon. Ranger 1 (P-32) and Ranger 2 (P-33) were launched to the Moon in 1961 by Atlas Agena Bs but both became stranded in parking orbit when the Agena stages failed to restart. Atlas Centaur 8 was launched on 8 April 1966 with a Surveyor development payload which was to be fired out of Earth orbit. This launch also failed to take the payload beyond the parking orbit.

Other National Booster Programmes

An interesting comparison can be made between the United States launch vehicle failure rates and those of the vehicle development programmes of other nations, not including theSoviet Union. Statistically, the most successful of these launch vehicle programmes was that of the French Diamant. Starting in 1965, that booster was fired 12 times with 10 successes for an 83% rate. The failures occurred with Diamant B-4 on 5 December 1971 when the second stage exploded at 147 seconds and Diamant B-5 on 18 May 1973 when the third stage failed to ignite. The Diamant was the first successful non Soviet-American booster.

The Japanese entered the launch vehicle field in 1966 with the Lamda 4S booster which was launched five times with four failures for a success rate of 20%. The Mu vehicle has been launched nine times since 1970 with only two failures for a 77% success rate. On 25 September 1970 and 4 February 1976 the Mu booster failed to orbit the intended payloads. The N launch vehicle was first fired in 1975 and, as of the end of 1978, was successful in four straight attempts.