Monday, August 1, 2005
Entry Into Eclipse
The airplane is made of conventional aluminum. The engines are small robust turbofans. And the controls are as standard as they come; cables, pulleys, and pushrods, no fly-by-wire. But dig a little deeper, and there is much more to Eclipse Aviation's Eclipse 500 very light jet, from an integrated electronics/avionics system to a unique method of airframe construction and a groundbreaking new small turbofan engine. Add to that Eclipse's early consideration of maintenance requirements and the result is the ingredients for an entirely new sub-category of general aviation jet.
Eclipse has firm orders for more than 2,200 Model 500s. The planned production rate, once the FAA grants type and production certification and the Albuquerque, New Mexico plant gets up to speed, is at least four airplanes per day and as many as six. At that rate, it won't take long before the skies are filled with Eclipse 500s. And all of those small jets, no matter how reliable, will need service.
To understand Eclipse's maintenance philosophy, it helps first to see why Eclipse even exists and what its founders are trying to accomplish.
"We're not a bunch of computer nerds who decided to build airplanes," said Michael McConnell, vice president sales & product support. "We're aviation people, aviation historians." By that, McConnell doesn't just mean that he and founder Vern Raburn, an early Microsoft millionaire, have long backgrounds in aviation--as do all of the Eclipse team members--but that what makes the Eclipse 500 the right airplane at the right time is an interesting confluence of events. It all started with airline deregulation in 1978.
"We think this was a watershed moment," said McConnell. At the time, production of general aviation aircraft reached an all-time high, nearly 18,000 aircraft in one year. Airlines were stable but costly and non-competitive, with prices determined by the government and little competition between carriers. Now, more than 25 years later, airlines are the new mass-transit, according to McConnell, and general aviation has become the more expensive option.
What Eclipse is attempting to do is change the value proposition for general aviation jet travel. The Eclipse people believe that current aircraft manufacturers are stuck on the same value curve, McConnell explained. The value enjoyed, or tolerated, by customers is distorted. Rarely, probably never, do customers find that the value of business jet travel improves while prices drop. To date, the benefits have increased, but so have prices.
At Eclipse, the microchip mantra of Moore's Law, which states that the processing power of computer chips will double approximately every 18 months, is also the foundation of the Eclipse value proposition. Or as McConnell said: "Increase capability at a lower price." And by lower price, Eclipse specifically means half the operating costs of today's entry-level business jet, a Cessna CitationJet. "If we can accomplish this," he added, "it will lead to market expansion. Value creates markets and value expands markets."
Of course, the rest of the aviation industry hasn't been sitting still, with competing new very light jets being announced with what seems like very high frequency. Amid the successes at Eclipse Aviation, there have been problems, such as a major program delay and redesign due to the initial engine selection of a Williams EJ22 and the switch to the Pratt & Whitney Canada PW610F. And recently, Eclipse raised the purchase price, admitting that the aircraft's material costs had increased beyond expectations from 2003. Eclipse remains on its current target for certification and first deliveries next year.
The way that everyone at Eclipse views the new 500 is not just as yet another general aviation jet but as a product that will disrupt the existing marketplace, just as the invention of the microchip or the Internet caused their own disruptions and created entirely new markets. And when markets get disrupted, the prognosticators are invariably hugely wrong in predicting the future. "Market forecasts for disruptive products are never correct," McConnell said.
No one knows how many very light jets will eventually be sold, and estimates have varied wildly, but the fact that Cessna has entered the market with its new Mustang and Embraer has also joined in with a PW600-powered design adds validity to the assumptions that gave birth to the Eclipse 500. Companies such as Adam Aircraft, Grob, Diamond, and others are also targeting the very light jet market.
While the questions surrounding the likelihood of the Eclipse 500 making it through certification have died off, a more focused critique has been, according to McConnell, the support issue. "Our competition," he said, "used to say the only way we can sell a business jet at this price is that the engines were not included, but leased. Today they are asking, `How will we support it?'" And that is a valid question for a company with no track record, no service centers, no product support network, no parts warehouses, and about a year to go until initial entry-into-service.
Design for maintainability
All aircraft manufacturers take maintenance into consideration, because customers demand predictable and reasonable maintenance costs. Maintenance man-hours per flight hour, in fact, have dropped since early generation business jets by a huge amount (see Chart 1, page 17), and the Eclipse 500 aims to continue that trend with an extremely low 0.5 maintenance manhours per flight hour.
The design of the Eclipse 500 is a large factor in helping keep maintenance costs down, but design can only go so far. There isn't a whole lot new in aircraft engineering, but there are ways to apply some intelligence to the process that will deliver the desired result: a light jet that can fly as much as 2,000 hours a year at extremely low operating costs.
Eclipse calls this intelligence "key technology enablers." The enablers include the engine, integrated avionics and systems, and innovative manufacturing technology. Eclipse's efforts to change the way manufacturing is done don't stop there, however.
"We're asking vendors to make a fundamental shift," said McConnell. "We're introducing a high-volume business model."
"This is by no means a slam-dunk for vendors," explained Dennis Whiteturkey, Albuquerque service center manager. Vendors are being asked to sign product-support agreements up front, to ensure that the path to keeping operators flying is clear from the beginning. The agreements aren't just for supplying material but also include some fairly stringent performance standards.
Eclipse asked brake manufacturer Parker Hannifin, for example, to try to bring per-landing costs down to less than one dollar. The initial figure was $2.64 and the number is already below $2, although it hasn't yet reached Eclipse's admittedly hard-to-reach goal. "These sorts of interactions are great," said Whiteturkey. "We're trying to drive costs down." While some vendors are still locked into the old ways of doing business, many are interested enough to work with Eclipse that they are willing to consider improved designs and technologies.
Early Eclipsers designed an airplane powered by two tiny Williams International EJ22 turbofans, which are so small that a technician could easily carry one under his arm. The EJ22 was supposed to be cheap enough that it might be simpler just to replace the engine at overhaul time.
After problems developed with the EJ22, Eclipse canceled its contract with Williams and eventually selected the PW610F, one of a family of new engines that Pratt & Whitney Canada had been working on. Cessna is also using a PW600-series engine in its new Mustang, which will also enter service in 2006.
The PW610F isn't a one-arm engine, but it will be easy to maintain. Initial targets are for a 1,750-hour hot-section inspection and 3,500-hour TBO. The hot-section can be done on-wing, with removal of just two of the four nacelle panels. Engine accessories are accessible by removing the lower nacelle panel. Engine change should take just 2.5 hours, and the engines can operate in either position with replacement of some parts that are specific to the left or right side. Dual-channel FADECs provide automatic fuel control. Pratt & Whitney Canada estimates maintenance costs for the PW610F to be dramatically lower than its PW200 or PW500 turbofans.
Avio-nics and electronics
The Eclipse designers are stretching the boundaries of general aviation jet design with the 500's integrated avionics system. The goal is for the system to manage all of the background activities of not just the avionics but also all the other aircraft systems. The pilot should receive information that is necessary to flying safely and efficiently, and when there is a problem that needs attention, the system will alert the pilot on one of the three large LCDs installed in the instrument panel. Trend monitoring and system health data, including problems that aren't important for the pilot to know about in flight, are saved for retrieval on the ground. Technicians can plug a computer into the system using common USB 2.0 ports, for both data retrieval and database and software updates.
Eclipse calls this system Avio Total Aircraft Integration. Avio isn't just about using technology because it's available, but more about applying technology to deliver results. In Eclipse's case, the desired result is more airplane availability and short turn times when maintenance is necessary.
Avio enables Eclipse to incorporate features that make the 500 more reliable and easier to maintain and at the same time, connects the avionics, electronics, and aircraft systems so they all work together.
Two redundant computers form the intelligence behind Avio. Communications within the avionics and aircraft systems take place on multiple redundant databuses so that there is always an available path for information flow, even if portions of some busses are compromised. This is a more network-centric approach to aircraft systems design, sort of like the fail-safe structures approach applied to electronic systems. Fail-safe design allows stress to take another safe path so as not to harm the overall structure. And in the Eclipse 500, busses communicate with each other to provide alternate routes for important information.
Eclipse's designers wanted to include as many systems as possible in the 500's integrated electronic backbone so they included system elements that make this possible. Instead of a bulky and complex hydraulic system with reservoirs, pumps, lines, and seals, the landing gear and flaps are electrically actuated using "smart" actuators. These actuators should greatly reduce maintenance costs because they are microprocessor-controlled and self-rigging. The designers use software to tell the actuators what to do and when to move, and a side benefit is that the software can set limits and regulate actuator acceleration to prevent overloading of pushrods. The software also allows the actuators to be self-rigging, which means that when a technician installs a new actuator, he or she does not need to perform any physical rigging but just checks that the software has made the correct adjustments.
The Eclipse 500 comes with built-in trend-monitoring software, which receives information from all the integrated components. Other components that help keep maintenance costs down include 127 electronic circuit breakers, which use internal microprocessors to react far faster to wiring faults than mechanical circuit breakers that depend on heating of wire to trip. Electronic circuit breakers also deliver useful health information on component output voltage and current.
To make use of all this information, technicians can download data from the onboard diagnostic storage unit. This unit stores aircraft data points along certain points of power-up, flight, and shut down, as well as continuous built-in trending data from every flight.
The trend in avionics design--computer-like line-replaceable units on easily removable circuit cards--is accelerating in the Eclipse 500. All avionics LRUs, the engine's FADECs, and critical parts of the electric power distribution system are mounted inside the pressure vessel to improve reliability.
The manufacturing process that Eclipse is using to build the 500 has received a lot of attention, specifically the friction-stir welding technique that zips together aluminum airframe parts. While it sounds all heart-warmingly high-techy, does friction-stir welding reduce maintenance costs?
It should. The process certainly is strong. During destructive tests, Eclipse has shown that friction-stir welded parts easily withstand forces that would pull rivets and promote cracking in fastener holes.
Friction-stir welding involves hugely expensive machines that press a mandrel onto two overlapping pieces of aluminum that are held into tooling fixtures by vacuum; the mandrel pressure creates heat that melds the aluminum together, leaving a smooth exterior surface bereft of the usual holes and fasteners. The process is extraordinarily quiet, making for a much more pleasant factory environment, and there are no smelly composites brews either because the airframe is primarily aluminum. Friction-stir welding is used for many parts, including frames, stringers, and doublers. Eclipse does use rivets and other fasteners, just not nearly as many as conventionally built airplanes.
Any Eclipse-size airplane typically takes 10,000 to 15,000 man-hours to build, according to Eclipse senior vice president and senior fellow Oliver Masefield. With friction-stir welding and a lean-designed factory floor, the Eclipse 500 will take fewer hours. But there's no getting around the cost. "That's prohibitive," he added. "You can either outsource it to a $10-an-hour country or do it intelligently." And this is key to not only the 500's low price but also the factory's ability to crank out significant numbers of the jet. Friction-stir welding and the optimized design elements cost a huge amount of money up front. Avionics development alone cost about $20 million and friction-stir welding another $25 million. The payoffs for those risks will be enormous because the "touch-time" needed to put together an Eclipse 500 will be minimized, and that will lower the cycle time for delivery of each airplane.
For Eclipse Aviation--and any manufacturer--what the customer experiences after delivery of the aircraft will impact the company's ultimate success. The 500 may be relatively inexpensive, but if it doesn't deliver on performance, reliability, and low cost of operation, it will never be successful. Jet buyers have more choice than ever these days and alternatives are available, although none (yet) meet the performance and cost promises offered by the Eclipse 500.
"You bought something," said Michael McConnell, "you just want it to work. We have to extend the value proposition beyond the purchase price."
Eclipse has been busy building an infrastructure to meet the needs of 500 owners and operators, including the much-anticipated high-volume air-taxi companies.
The MSG-3 (maintenance steering group, 3rd generation) concept was an importance consideration during the 500's design and will remain a focus as the jet gains field experience. A manufacturer participating in an MSG-3 process analyzes the data from the field to learn how to improve maintainability.
Structurally, the 500 is simply an aluminum airplane with fewer rivets and fasteners than a conventional jet. Airframe repairs can be done using riveted aluminum patches; there will be no friction-stir field repairs. For large damaged parts, replacement will be more cost-effective than repair. Major structural elements like a wing, for example, are considered LRUs and will be easily replaceable.
In considering maintainability, Eclipse engineers wanted components to be measured by their mean time between unscheduled removal rather than mean time between failure. The removal figure is more accurate because it takes into account no fault found problems where testing of a removed LRU shows that there is nothing wrong with that LRU. By asking component suppliers to deliver a high target for mean time between unscheduled removal, Eclipse has set the bar higher and the result should be greater reliability and lower costs.
Other maintenance features include visible wear indicators on the brakes, a 20,000-hour (initial) airframe life, and proximity switches to measure position of landing gear, flaps, and other moving parts instead of error-prone and unreliable microswitches. Interior lights are LED instead of incandescent; LEDs last for hundreds of thousands of hours and generate hardly any heat. Exterior position and anti-collision lights are also LED, while landing lights are high-intensity discharge types and wing-ice inspection lights and taxi/recognition lights are halogen. The interior can be installed in just one hour so that won't be a burden for maintenance crews. And something that ought to be simple, a windshield change, will be simple, with just a rubber gasket and no need to first remove the instrument panel. The goal is for technicians to be able to swap out both windshields in just 10 hours, plus a four-hour aero-seal cure time, a tenth of the time needed for other small jets.
Finally, there are only six life-limited parts on the Eclipse: the ELT battery (24 months); secondary outflow valve filter (1,200 hours); engine shock mounts (engine TBO); crew oxygen regulator (overhaul at six years); air-conditioning compressor belt (10,000 hours); and oxygen bottle (15 years).
A manufacturer can opt to create an authorized service center network, its own service centers, or some combination of those, with varying levels of service provided by vendors of components, engines, avionics, etc. The resulting support scheme has enormous implications for the customer.
"We have to extend the value proposition beyond the purchase price," said McConnell. He is adamant that customers not have to deal with multiple vendors for service and warranty, and the Eclipse support model is built around a single Eclipse point of contact for any service issue. "We convened five different focus groups," he added, "and they loathe and despise having to call everyone for information." Or, as Bill Caswell, director, customer products & services, explained, it boils down to an argument over "whose electrons are messing up? There's no way customers will tolerate that sort of latency."
McConnell and crew spent a lot of time analyzing the best way to deliver maintenance to Eclipse customers and found that the traditional retail business jet maintainers are focused on low-volume, high-dollar jobs and are not structured to deal with the big flow of airplanes that Eclipse might bring to their hangars. Consequently, Eclipse is building its own service center network in the U.S., but will work with authorized service facilities for non-U.S. customers. For the flight department that employs its own technicians or small maintenance shops, Eclipse will offer the opportunity to have technicians trained on the Eclipse 500 so that the owner can be reimbursed for warranty claims. These technicians will be designated as Eclipse authorized repair technicians.
Eclipse plans to offer maintenance cost-per-hour programs with guaranteed dispatch reliability and direct operating costs for fleet operators. For a large-enough fleet, Eclipse might build a fleet service center at an operator's headquarters. The operator would be responsible for line maintenance while the local Eclipse team would handle all other maintenance.
To meet the needs of the 260 Eclipse 500s scheduled for delivery in the first year of production and the 880 during the second, Eclipse will build seven Eclipse-owned service centers in the U.S. and smaller satellite facilities near large fleet operators. The first facility is already under construction, at Eclipse headquarters in Albuquerque, which will also include a parts depot. The second center is in Gainesville, Florida, and this will be followed by one in the Northeast. The first two will be ready by the time Eclipses delivers 100 airplanes in 2006, and all seven will be open for business by 2008.
Each facility will be designed for optimum efficiency, with a similar layout to an automobile dealer's service bays. The 45,000 square foot hangar, with doors on each end, will be able to fit two rows of six Eclipse 500s, one row parked tail-in on each side of the hangar, leaving a center aisle free. This eliminates the need to move one or more airplanes to get another out of the hangar.
In designing the new service centers, Eclipse experts have evaluated the entire customer experience and the performance of maintenance tasks to find out how to deliver better service. McConnell is enamored of the maniacal customer focus he witnessed while visiting Lexus automobile headquarters in Torrance, California and doesn't see why that business model can't be applied to aviation. Carefully crafted business processes and a focus on customer satisfaction will be the rule. "Our focus will be on the customer," said McConnell. "What does it take to get him back in the air? We're going to do the right thing."