The Ariane 5 cryogenic propulsion system consists of the main engine, feed lines, valves, pneumatics and tank-pressurisation systems and is the responsibility of France's Soci,t, Europ,en de Propulsion (SEP) which, as a prime contractor, leads a group of 37 European companies

"Our objective was to develop an extremely reliable low-cost engine, and not to achieve the highest possible performance," says SEP's Vulcain programme manager, Serge Eury. "If we lost some performance because of our technical choices, we compensated by increasing dimensions - a bit more propellant in the tanks is worth it if we gain the extra reliability."

This philosophy drove the choice of a gas-generator cycle for the Vulcain, instead of the topping cycle used in the Shuttle main engine. Gas generation, in which a gas generator drives the liquid- oxygen and liquid-hydrogen turbopumps which feed fuel to the main engine, is already used in the Ariane 4's SEP-built cryogenic HM7 third-stage engine. Its main advantage, says Eury, is that the engine works under around half the pressure in the main-thrust chamber. "We lose about 15s of specific impulse, which translates to about 900kg to GTO," he adds. Studies of both cycles revealed that the cost of developing the gas-generator engine was "about half as much", however, while production costs could be reduced by 30%.

The Vulcain produces 1,120kN thrust, and burns for around 600s, ejecting some 250kg/s of gases at 787,000ft/min (4,000m/s) through the nozzle. First tested in April 1990, the engine has since built up more than 66,000s running time, and will have amassed 85,000s by the end of the year.

Starting is provided by a solid charge-starter, which spins up the turbo-pumps and by pyrotechnic igniters, which light the gas generator and main chamber. The gas generator is fed by propellants, taken from the main fuel supply, its exit gases, powering the liquid-hydrogen turbo-pump at 33,500RPM and its liquid-oxygen counterpart at 13,800RPM.

Problems experienced with the HM7 are not likely to appear on the Vulcain, says Eury. "The design is completely different," he says. "In the HM7, the turbo-pumps were connected mechanically and geared to each other. In the Vulcain, they are separate, which means we can optimise the speed of each and mount them symmetrically about the engine, reducing thermal deformation."

SEP has also done away with the complex dynamic seals used in some turbo-pumps, which are both expensive to manufacture and, while yielding better performance (through reduced leakage), can be less reliable. Instead, floating ring- seals between the two sides of the turbo-pumps are employed, with helium passed through the seals to isolate the two halves.

Turbo-pump tests have gone well, the only setback being the explosion of an oxygen turbo-pump in 1994. This was a refurbished pump, however, which did not incorporate design changes decided in 1990. The subsequent investigation, which took four months, identified nine possible causes of the failure and subsequent fire. As a result, the turbo-pump casing material has been changed from aluminium to inconel steel, the size of the turbine bearings has been increased and other alterations have been carried out.

Testing of the main engine and associated system has been exhaustive. A total of 14 engines have been used in the test programme, five of which have been run for more than 6,000s. A further two engines will be added to the programme in July for the final flight-qualification campaign. This has been delayed following the premature shutdown (because of a programming error) during what was supposed to be the first full 600s firing.

This has been delayed slightly, following the premature shutdown (because of a programming error) during what was supposed to be the first full 600s firing.

Source: Flight International