The Ariane-5 Cryogenic Main-Stage Development Tests

J. Gigou

Directorate of Launchers, ESA, Paris

J.Y. Heloret

Cryogenic Main Stage Manager, Aerospatiale, Les Mureaux, France

The Ariane-5 launcher development programme was given the go-ahead at the Meeting of the ESA Council at Ministerial Level in The Hague in September 1987. Development of the cryogenic main stage, the launcher's core stage, represented a significant shift in scale compared with that of the cryogenic stages previously developed in Europe: the H8 and the H10. One of the key milestones in developing the new stage was the test firings at the ELA-3 launch facilities in Kourou, French Guiana. The data from the first 37 seconds of Ariane flight 501 on 4 June 1996 have been analysed to determine how the stage behaved in flight.

Figure 1. Lift-off of Ariane flight 501 on 4 June 1996

The context and organisation of the tests

After various studies and deliberations, it was decided to divide the cryogenic-stage test firings into three main campaigns:

• A battleship (BS) test campaign, using a dedicated thrust-frame and engine representative of the flight configuration, and industrially produced tanks to some extent functionally representative of flight tanks but incorporating wide structural margins (dual-walled stainless-steel cylindrical sections and separate bulkheads). The functional electrical equipment was housed together in an electrical bay providing easy access.

• A development (M) test campaign, using a stage representative of a flight stage apart from a few details: no safety subsystem, no passivation or neutralisation system, since these items would contribute nothing to the tests and might even cause serious problems.

• A qualification (Q) test campaign, in a configuration similar to the M tests, but using flight-standard equipment.

Two major decisions were taken in 1989 concerning these tests:

• All testing would be done at ELA-3, so as to test stage operations in a real environment at the earliest opportunity.

• M and Q testing would be done on the same stage, except that any equipment whose internal definition changed significantly would be replaced. This decision had been necessary for budgetary reasons.

Taking into account the location of the testing and sequence of objectives - in particular, that the prime objective of the battleship tests was to ensure that the propulsion system functioned properly - the organisational arrangements were as follows:

 - Prime contractors for tests:
   BS and M:                    SEP
   Q:                           Aerospatiale

 - Stage prime contractor:    Aerospatiale

 - Tests conducted by:        CNES ground
                             subdirectorate

The test campaigns in Kourou

The battleship campaign

Figure 2. The battleship (BS) and M stages

Figure 3. The battleship stage under test

The battleship campaign started on 2 September 1994 with the test review. This important meeting, at which all parties involved in the main-stage programme examined the preparation status of the specimen and test facilities, gave the go-ahead to start the campaign. The first test (BS1.1) subsequently took place on 23 September, but, due to a number of problems, first-stage ignition did not take place until 17 November during BS2.2. This test completed an arduous phase which made particularly heavy demands on both the ground and launcher teams. The outcome, however, was an outstanding success, with 281 seconds of perfect operation, proceeding entirely as planned.

What had happened between BS 1.1 and B52.2? The answer is simple: for the first time, a stage and all associated ground facilities, including all hardware/software features, had to be operated simultaneously. Various problems arose and these quite naturally gave rise to procedural changes, which had to be fully validated before operational application. In addition, the engine's launcher-side hydrogen feed valve failed to open and had to be replaced. Thereafter, the campaign proceeded without any major anomalies.

The last test took place on 27 January 1995, completing a fine run of successes, namely four long-duration test firings without any serious problems coming to light.

The successful battleship campaign thus validated stage operations from both the onboard and associated ground facilities viewpoints. It validated the ground safety systems, thus allowing a stage (the MQ - see below) to be set up for testing in its flight configuration. It also provided confirmation that the stage preparation teams were able to carry out operations to tight deadlines, while keeping to the stringent quality assurance rules that govern all our activities. Last but not least, the campaign verified that the stage design was functionally viable and reliable.

The development campaign
The development (M) campaign started on 20 April 1995 with the test review. The first test was delayed by a serious accident in which two technicians sadly lost their lives, a tragedy that deeply shook the entire space community.

The first test took place on 22 May. It did not get as far as the test firing, essentially due to ground operations software malfunctions in the real environment. The following test scheduled for 30 May was halted during the engine start- up sequence by an incorrect parameter specification in the flight program. The next test took place successfully on 16 June. The stage operated perfectly for the planned firing duration of 589 seconds.

The campaign then entered a difficult phase. The ground operations software was still undergoing modifications to handle the 501 launcher after the main-stage test phases. Some of these modifications were integrated into the operational system, and rightly so, since it is always very worthwhile testing the latest version of a system at the earliest possible point - especially one as complex as a synchronised launch sequence. The main-stage M test campaign therefore had to allow for the secondary objective of validation of the full-scale ground/onboard/software system. Thus, the M2 test on 3 July did not go through to engine ignition. After a break in the campaign, which was used to carry out further validation, the M3 test likewise failed to produce a successful conclusion. Two major incidents required fairly substantial work on the stage: an oxygen leak detected during M3 had to be sealed, which could only be done from inside the tank itself, and a burst high-pressure oil line feeding the flight control actuator system called for an intricate repair operation to replace the affected equipment.

Testing was then able to resume and the M3.2 test took place on 23 October. During the 620 seconds of burn time, the stage functioned wholly satisfactorily. Next came the M4 test (592 seconds) on 7 November, followed on 23 November by the M5 test (623 seconds), the last in the campaign.

A number of incidents and accidents had punctuated the campaign, reinforcing the belief that ground/launcher system operations of this type are particularly complex. This complexity, which is only to be expected of a system of this scale, is such that any modification, however innocuous it might appear, has to undergo very extensive validation phases. The campaign also showed that for activities such as ours in which all systems are interdependent, no simulation will replace actual-scale testing, which will always be essential. Last but not least it showed that test incidents - which are not necessarily abnormal (testing would not be necessary if it was absolutely certain that there was nothing to learn!) - play an important part in securing system reliability.

The qualification campaign
The qualification campaign started on 1 December 1995 with the test review and lasted just over a month. It actually involved two tests: the Q1 test which took place on 15 December and lasted 628 seconds: and the Q2 test of 594 seconds duration, which took place on 6 January 1996.

The Q2 target H0 was the same time as the flight 501 H0. Apart from the stage test objectives, Q2 also helped demonstrate that the launch teams could handle a long countdown, under the same conditions as the flight 501 countdown would occur. Q2 thus completed the lengthy phase of main-stage testing in French Guiana. The MQ stage used for the M and then the Q tests - after replacement of some equipment - clocked up over 3600 seconds of running time, i.e. about six times its nominal burn time.

Flight 501
The 501 main-stage ignition sequence proceeded perfectly and electrical system operations were wholly satisfactory. As far as the pogo phenomenon was concerned, vibration levels were close to those recorded during the development (M) tests. Large and constantly increasing pressure fluctuations in the actuators were detected as from 20 seconds onwards, due to a phenomenon known as buffeting, but this does not cast doubt on the functioning of the stage. Overall, the main parameters of the command system were close to those forecast. The thermal environment in the stage bay was cooler than during M testing, but this had no impact on overall equipment functioning. The Vulcain engine functioned perfectly.

Figure 4a,b. The M stage and its testing

Conclusion

The cryogenic main-stage test firings were carried out over a period of almost a year and a half; between September 1994 and January 1995. These tests revealed some difficulties in mastering stage operations, primarily at the start of the battleship campaign and during the development test campaign. More specifically, most of the difficulties arose during the synchronised sequence - the final phase of the countdown involving combined electrical and fluid commands.

Onboard functioning did not reveal major difficulties; a few incidents such as the liquid-oxygen line feedthrough leak or a burst in an engine actuator unit line disrupted the development test campaign, but these problems were overcome and corrected without any great impact on overall scheduling.

Finally, it should be noted that the decision to conduct the stage testing at ELA-3 was vindicated by the Ariane flight 501 final countdown to lift-off, both of which proceeded very well and without interruption, apart from weather-related problems.