As airliners evolve, so do their auxiliary power units.
Andrew Doyle/LONDON
AUXILIARY-POWER-UNIT (APU) manufacturers are employing new technology to meet ever-increasing demands from aircraft operators for higher reliability and low cost of ownership. They are also studying radical new roles for the APU, and the likely requirements of the new large airliners under development by Boeing and Airbus.
The main functions of an aircraft APU are typically to provide power to start the main engines, pneumatic and electrical power for ground operations, and back-up electrical and pneumatic power in-flight. The manufacturers, however, are increasingly taking on responsibility for the whole APU installation, including mountings, the fire-proof enclosure, fire detection, generators, air inlets, exhaust and other ducting, and even the relevant control panels in the cockpit.
Another aspect is the drive to simplify the design of APUs, by cutting the number of moving parts where possible. This may culminate with the adoption of so-called more-electric technology, where all of the aircraft's systems are electrically powered. This would eliminate the need for driveshafts and gearboxes around the APU, with a generator on the shaft of the engine used instead.
AlliedSignal Engines, the world's largest APU producer, manufactures 20 basic models, with a total of 41 variants. Five of those models have been developed during the past three years: the 331-500B for the Boeing 777; the 131-9D for the McDonnell Douglas (MDC) MD-90; the 131-9B for the 737-600/700/800; the RE220 for the Gulfstream V and Bombardier Canadair Global Express; and the RE100 for the Learjet 45 and Cessna Citation Excel business jets.
The 131-9B for Boeing's new-generation 737 will be the first in the world to incorporate a so-called starter-generator, which eliminates the traditional DC starter motor and associated clutch and gearbox, and the need for a dedicated APU battery. AlliedSignal was picked by Boeing as the sole-source supplier for this aircraft.
At the heart of the system are the start-power unit (SPU) and the start converter unit (SCU). The SPU converts AC power from a ground-based supply (when available), or battery power, to 270V DC. The SCU converts this to regulated variable-frequency power which causes the generator to perform as a starter-motor. The whole process is controlled by the APU's electronic-control unit.
According to Rick Berry, AlliedSignal's engineering manager for new business for commercial APUs, the technology has been under development for the past five years. "It improves APU starting by providing a smoother continuous torque input," he says. "It can start the APU from the completely cold-soak condition with low [volumes of] ram-air."
The 131-9B is scheduled to receive its US Federal Aviation Administration technical standard order on 15 July, followed by delivery to Boeing of the first production unit on 26 August. The first flight aboard a 737-700 is due in February 1997.
Another of the innovations recently introduced by AlliedSignal is known as eductor cooling, which was first incorporated in the 777's 331-500 APU, and is also included in the 131-9B and RE220. The eductor nozzle replaces the high-power cooling fan which was previously used in the oil cooler.
By eliminating the cooling fan, AlliedSignal claims that reliability is improved because of the reduction in the number of moving parts, and says that starting performance is improved because fan drag is also eliminated.
The oil cooler also tended to become clogged by leakage from the fan seal. According to Tom Johnson, vice-president, commercial APUs at AlliedSignal, the eductor design was customer driven and "-grew out of design reviews with customers early in the 777 programme".
Auxiliary Power International (APIC), a joint venture between Sundstrand of the USA and France's Turbom,ca, is also studying eductor technology, but is unlikely to introduce it until it begins developing an all-new APU. According to APIC manager of corporate marketing Karl Johanson, this is because the company is satisfied with the reliability of its existing cooling fans.
RELIABLE COOLING
The company manufactures the APS 2000 for the Boeing 737-300/400/500 and the APS 3200 for the Airbus A319/320/321, and is developing the APS 2100 for the MDC MD-95.
"The APS 2000 cooling fan has a mean time between unplanned removal [MTBUR] of over 225,000h," says Johanson. "There is a trade-off that is not yet really apparent to us. It's pushing the cost from a rotating part into a fairly expensive static part." The eductor technology is difficult to apply to existing aircraft because "-you have to alter the aerodynamics and structure to accommodate the air inlet", he adds.
Although APIC is now only in the narrowbody market, it plans to expand its range upwards with products based around Turbomeca's existing line of gas generators. Not included in the joint venture is Sundstrand's APS 500, used on the Bombardier de Havilland Dash 8-100/200/300 and several business jets and helicopters, and the APS 1000, installed on the Avro RJ and some British Aerospace 146s, along with the Saab 2000, Fokker 50 and IPTN N250 turboprops.
Johanson says that APIC is also not convinced of the merits of the starter-generator, although he accepts that the technology makes sense for the new-generation 737. "There are some significant cost drivers there that cause the starter-generator to have a difficult time justifying itself on the aircraft," he says. "It's very expensive."
APIC also proposed a starter-generator when it tendered for the new-generation 737, however, because it is seen as "a unique case". Both it and AlliedSignal proposed much larger APUs than are used on the existing 737, and this would have required the addition of a second, dedicated, APU battery.
By contrast, says Johanson, a starter-generator was never considered for the MD-95's APS-2100 because "-if it doesn't have a need for a second battery, it doesn't buy its way on to the aircraft". He adds that starter-generators are unlikely to be used on new APUs under development for the proposed new generation of 100-seat regional jets.
The elimination of the external gearbox, meanwhile, looks unlikely to be achieved in the near future, although current studies are "-advancing into the demonstration phase", says AlliedSignal's Berry. "We're looking for targets of opportunity." Success would probably yield significant improvements in reliability, by eliminating a large number of moving parts.
The concept, which is also under study by aero-engine manufacturers, is part of a wider thrust towards the more-electric aircraft. It involves the replacement of the external gearbox and generator with a switched-reluctance generator located on the shaft of the engine.
"We have offered designs to Boeing that integrate the generator into the gearbox, which we see as a first step towards a rotor-mounted generator," says APIC's Johanson.
The main problem with rotor-mounted generators is that they would have to run at high, variable speeds - up to around 45,000RPM. Existing generators are geared to run at much lower, constant, speeds - 24,000RPM for the APS 3200, while the 131-9B will run at 12,000RPM. Most electrical equipment aboard airliners which have more than 100 seats is designed to use a constant 400Hz supply.
"The real driver [for rotor-mounted generators] will be if all aircraft systems are redesigned to operate with variable-frequency power," says Johanson. "It's happening in the military world and it will migrate to commercial aircraft."
The use of new materials also has the potential to boost the reliability and performance of APUs, AlliedSignal believes. It is conducting a trial with Alaska Airlines to test ceramic static vanes in the turbine-nozzle sections of eight of its 85-98DHF APUs, fitted to MDC MD-80s. The lead unit has accumulated over 3,000h with the ceramic vanes, and the US carrier is "very happy with the results", says Johnson.
The ceramic vanes are designed to be more resistant to oxidisation, corrosion and erosion (or "sandblasting"), he adds. They could also be used to allow an increase in turbine-inlet temperatures, boosting the cycle-efficiency of the APU. They are, however, more expensive to make than traditional metal blades.
"We're at the stage where we're evaluating the Alaska Airlines results," says Johnson. "If we feel that the airlines are really gaining a benefit here, it looks likely that we will be advancing this across the fleet in the next six months."
"There are great hopes for ceramics," agrees APIC's Johanson, "and the most immediate opportunity is in the nozzle vanes. Their use can eliminate the need to give up bleed air to cool the turbine nozzle material. However, APIC is not there yet because we're not pushing our temperatures that high." It is likely that ceramics will eventually be considered for both the static and rotating sections of the turbine, he adds.
OPERATION IN FLIGHT
Other studies are focused on improving the reliability and reducing the weight of the electronic-control units with which the vast majority of production APUs are equipped, and the development of so-called "oilless bearings".
It is not only new technology, however, which has the potential to improve the overall effectiveness of APUs. Studies are being carried out on radical changes in the way that APUs are used, to allow the aircraft's main engines to be operated more efficiently.
One example is the possibility of having the APU operating continuously during flight, to improve the fuel consumption of the main engines. Bleed air from the main engines is used to provide air conditioning during flight, and also drives generators through external gearboxes to provide electrical power.
As a consequence of these demands, the engines also develop less power than if they were "clean". The aircraft's systems also have to be laid out in a relatively complex way, because of the location of the engines.
AlliedSignal is studying with aircraft manufacturers the concept of continuous APU operation. "We have been evaluating this with an advanced design group at Airbus," says Berry. "It looks like there are some benefits," he adds, although the "-jury is still out" on the long-term gains.
The most likely project which could make use of a continuously operated APU is the Airbus A3XX, as Boeing's proposed 747-500X/600X is a derivative aircraft which will probably be fitted with a more powerful, but conventional, APU. AlliedSignal and Pratt & Whitney Canada are expected to bid for the A3XX business, against APIC's so-called APS 4XXX. The stretched 747 competition is expected to involve AlliedSignal and P&WC, and result in another single-source contract.
Such would be the demands of an aircraft the size of the A3XX that another option under consideration is for the aircraft to have two APUs. This would be required to provide redundancy if the role of the APU is expanded to include in-flight operation.
DUAL APUS UNDER STUDY
Initial studies are focusing on whether the reduction in fuel consumption, resulting from less energy being bled from the main engines, is sufficient to justify installing much larger, heavier, APUs. If dual APUs are used, airlines would be likely to insist that the aircraft be capable of being operated with only one unit serviceable, to maintain a reasonable dispatch reliability.
APIC's Johanson believes that such large APUs would not present problems from a technical point of view, and that having two of them could make sense. "Redundant APUs are not an unacceptable concept," he says. "There is no reason that dual APUs will not be found on some of the large transports. As long as you can get away with one, you're better off having redundancy."
According to Andrew Cion, market account manager for APU products at P&WC, the Canadian company will probably make "as few changes as possible" to its PW901A APU for the A3XX and 747 stretch competitions, although the final requirements for both types are not yet fully defined.
P&WC is the sole-source supplier (with the PW901A) on the 747-400, its only APU product to date, with more than 350 deliveries so far. Cion says that the company has no firm plans to expand its APU line, although "-we're always reviewing market opportunities".
A growth PW901A would probably incorporate "hot-end material changes", says Cion, using current-generation, single-crystal, turbine-blade technology, rather than ceramics. There will also be a redesigned compressor, "-depending on the growth required".
The emphasis will be on maintaining the reliability levels of the PW901A, and retaining substantial commonality, as most, if not all, of the 747 stretch customers will already operate large fleets of 747-400s. P&WC claims that a "schedule interruption rate" (or delay in excess of 15min) is attributable to the APU of one per 4,975 departures on the -400.
While Cion acknowledges the operational benefits of dual APUs, he argues that it may not be justified on cost grounds. "The only drawback is the cost of ownership," he says. "If you can get away with one, that always makes more economic sense."
Manufacturing and assembly costs are being addressed too, through the adoption of so-called integrated product development. According to AlliedSignal's Johnson, the company "-had responsibility for designing the entire installation" on its latest models, an approach which he believes to be "the way of the future". APIC's Johanson agrees that this is "very much the current trend".
"On the MD-95, we own pretty much everything that touches the APU, right out to the skin," says Johanson. "As the airframe manufacturers look for more efficient ways to build aircraft, it is more logical to have the APU manufacturer design the installation.
"On the production line, the further the aircraft can go down the line before the assembly is incorporated, the fewer the man-hours that are being incurred by the manufacturer. MDC made a concerted effort to examine that."
APU manufacturers are also co-operating much more with the suppliers of equipment which interfaces directly with the APU, such as the devices that start the main engines.
IMPROVING RELIABILITY
"The main engines require our bleed air to start," says Johanson, "so they want a starter that's optimum for the power being supplied to them. If we do that in harmony [with the supplier] you get the best system, and a higher probability that it will work first time."
Meanwhile, the APU manufacturers are working to improve the reliability of existing models. AlliedSignal's latest figures show that overall reliability levels of its in-service commercial APUs, in terms of MTBURs, declined slightly between the beginning and the third quarter of 1995, although the MTBUR has risen by 80% over the past six years. The company says that the decline is mainly attributable to the increasing average age of the installed base.
"There have been some failures earlier than anticipated," says AlliedSignal's Johnson. "We have developed some service bulletins to improve reliability on those APUs. There is not a significant reliability issue now."
For the 85 Series a new manufacturing process for the turbine disc has been introduced, and field evaluations of a new compressor seal are due to be completed by the third quarter of this year. A new second-stage compressor is also being introduced. A new compressor seal is being incorporated in the 331-200. For the 700-4E, modifications to rectify high oil consumption were made available earlier this year.
Whatever the final configuration of the new generation of large airliners, it is clear that the new generation of APUs will have a crucial role to play.
Source: Flight International
