Exactitude in manufacturing should drive down MRJ costs
The card that every new commercial aircraft should get to play is a decisive rise in operating efficiency. Being new, the airliner should offer fresh technology that lets it transport passengers for less expense. For a mature rival program, a card to play in reply is low manufacturing cost. Having slid well down the learning curve, it can, if necessary, be priced more cheaply than the fancy new competitor.
It is not working out quite like that in the regional jet market, however. Because of more than three years of development delays and a vigorous technological response by an established competitor, Mitsubishi Aircraft's forthcomingmay quickly lose much of its efficiency advantage. First delivery for the MRJ program is now due in mid 2017, but by 2020 a heavily updated version of the most closely comparable E Jet should go into service. The -E2, as it is called, will reply not only by adopting the MRJ's engine; but it will also have improved avionics and, most notably, a new wing (AW&ST Sept. 23, p. 18).
But it turns out that Mitsubishi Aircraft has another card up its sleeve. The company says the MRJ has been designed to very high levels of precision, with the aim of easing assembly and thereby cutting production costs from the very beginning of the manufacturing program. No cost figures are available, but evidence of high-quality manufacturing was apparent when Aviation Week last month saw fuselage major assemblies of the first prototype, built by airframe contractor. Other cost savers will be a moving assembly line and oven-cured composite parts.
MRJ development costs, meanwhile, are rising by about ¥20 billion ($200 million) in engineering salaries for each year of delay, says Mitsubishi Aircraft President Teruaki Kawai. Since the original budget for a five-year, nine-month program with entry into service at the end of 2013 was ¥150 billion, development now looks like it will cost ¥220 billion, though Kawai says that can be easily absorbed in a program with decades of production life.
Production engineers at MHI are making use of the extra time by driving more risk out of fabrication and assembly processes—probably not a small benefit for a manufacturer that is not skilled in making all parts of commercial aircraft. The program can hardly afford manufacturing foul-ups after suffering the development delays, which were mostly caused by Mitsubishi Aircraft's lack of experience in airworthiness certification (AW&ST Sept. 9, p. 50). Three customers are awaiting delivery of 165 aircraft from the program, which so far is developing the MRJ70 of 76 seats and MRJ90 of 92 seats, the latter attracting the most demand.
From the outset, the program envisaged high costs for design and early fabrication of detail parts in return for permanent savings in assembly. The idea is simply to design and build the aircraft with tighter tolerances than are normal in commercial aircraft so that parts fit together better, greatly reducing the time taken by adjustment, such as shimming. The tolerances of detail parts in the MRJ are typically only half of those normally used in other manufacturers' programs, says Hiroyuki Tatsuoka, deputy director of MRJ manufacturing assembly at MHI, declining to give details.
As production proceeds, the cost of fabricating highly precise parts will quickly fall to normal levels, says Kawai; making them does not take more machining time than less-exact components do. Rather, the main cost has been the greater engineering effort in designing the aircraft for higher precision. But that is a once-only burden on the program, whereas quick and easy assembly will be a cost saver until the last MRJ or MRJ derivative is built. The benefits are not just in final assembly, usually thought to account for about 5% of commercial aircraft costs, but also in making the major assemblies, such as fuselage modules, and also subassemblies.
Each of the fuselage modules of the first prototype took about half as much time to put together as might have been expected with normal fabrication precision, Tatsuoka says. Thanks partly to prior practice on engineering prototypes, no major problems arose, and almost no adjustment was needed to make parts fit. Making shims—inserts that fill unintended gaps between parts—is a seriously time-consuming process in assembly, because measurements must be made for each one, which is then customized for its joint.
Normally, about 10% of holes will require reaming to get a fit when pieces are joined for riveting, laying one hole over another, Tatsuoka says, without specifying which parts of the aircraft that ratio applies to. For the MRJ the fraction was much less, “a very small number.” The tolerance of drilled holes is normal, however, because MHI is using standard rivets, says Akihiko Ishikawa, general manager of MHI's MRJ manufacturing program.
Availability of precise manufacturing machinery appears to have been a key enabler for this design and production strategy. MHI, Mitsubishi Aircraft's majority owner, needed only to buy few machines to build the MRJ's structure; its fabrication equipment could already achieve the necessary tolerances. In earlier programs, the precision capability had not been fully used, says Kawai. That may seem surprising, but he adds: “Almost all the machines in Japan are precise.”
Tight tolerances should also pay off in MRJ operation. Carefully manufactured skins with a surface finish within 10 microns (0.4/1000 in.) are mounted on precisely built frames, minimizing drag. The smoothness of the skin of the first MRJ prototype, the tight alignment of its panel joins and the regularity of its riveting were readily apparent during Aviation Week's visit to the Tobishima factory in southern Nagoya.
“The assemblies look very clean for early articles,” says an experienced production engineer from another manufacturer, examining photographs of the prototype. “There are no apparent tool marks [and] the fasteners are all the same size and in a consistent pattern, so it does not appear that there's been workmanship problems or even rework caused by an engineering issue or change.”
Visible fasteners on the MRJ prototype all appear to be seated properly, meaning there are no gaps underneath the head or collar. Accurate drilling has evidently kept them perpendicular to the surface, as they should be. From the outside, at least, sealing between panels looks like high-quality work.
A consultant production engineer who has also examined close-up photographs agrees that MHI has done a good job on the prototype.
The global aerospace industry is increasingly adopting moving assembly lines, in which the product continuously edges through the factory, creating a production beat and reducing wasted time and inventories. Since this is part of the lean manufacturing philosophy, largely learned from Japan and especially Toyota Motor, it is not surprising that MHI is applying the technique to MRJ production. The company has hired the Japanese consultant who advisedon the process, says Kawai. MHI instituted its first moving line at Nagoya and then applied the process to production at its factory in Hanoi, where it makes flaps for .
Although the MRJ is largely built of conventional aerospace aluminum, carbon-fiber reinforced plastic will be used for the moveable surfaces and for the tail. For the latter, MHI is using a process called A-Vartm—a variation of vacuum-assisted resin transfer molding. In this process, carbon fibers are infused with resin partly by suction, so the cost of an autoclave is unnecessary. MHI and material supplier Toray developed the A-Vartm variation.
Two other cost savers for MRJ production are the adoption of a circular body cross section, which has simplified tool design, and the use of conventional aerospace aluminum, that is cheaper than so-called advanced aluminums and aluminum-lithium alloy. Those newer metals offer weight savings that Mitsubishi Aircraft judged unjustified for a regional aircraft. A 2009 decision to build the wing of aluminum, instead of carbon-fiber composite, also saved money, though the main aim was reducing weight for the MRJ70.
Of MHI's plants around Nagoya, the one at Oye is making detail parts and the tail surface assemblies while Tobishima is building the major assemblies, including the outer wings and the fuselage modules, which it is also joining to create the whole fuselage, incorporating a wing center box. This includes fitting onboard equipment. The Komaki South plant, also at Nagoya, will perform final assembly, checking out and flight testing. MHI is building six MRJ airframes, five for flight-test aircraft and one for static ground testing. Pratt & Whitney will supply PW1217G engines.
In September, the fuselage sections of the first prototype were lined up but not connected. The outer wings were still being made, so a manufacturing prototype of the left outer wing was placed alongside them to show the final appearance. MHI also displayed a completed left winglet. The machinery for series production was employed in making the parts for the prototype.
Approaching an aircraft manufacturing program, production engineers first address the processes and parts where they see the greatest risks. For the MRJ, MHI engineers focused particularly on the wing, where the challenge to precision was the length of spars and skins, and the nose and tail fuselage sections, which presented the difficulties of double curvatures. Those end sections also typically use thicker skins, adding to the challenge in achieving required tolerances for length, width, curvature and so on.
The nose, especially, was seen as a risk, because MHI had no experience in making that part of a jet transport, says Tatsuoka. “We had to make it very tight, but for us, it was new,” he says. “We tried many methods to do that, starting with windscreen,” around which the rest of the nose was built. “There were several different ways to do that.” The outer wing boxes, too, required repeated attempts.
Development delays have given the production engineers time to progressively address smaller risks. Tatsuoka gives the example of drilling vertical holes in the upper chord of the center wing box, where that part protrudes for joining with the skin of the outer wing boxes. It is a classic problem area in commercial transport aircraft, because of the constricted space beneath the fuselage. Tatsuoka says the team has now worked out how to drill the holes there properly. It was an issue that would have been worked out much later in the manufacturing development schedule had the aircraft made its first flight in late 2011, as initially planned.
Mitsubishi Aircraft says that, with two big delays caused by its unfamiliarity with airworthiness certification now behind the MRJ, design development should proceed normally. If the production engineers have used the unexpected time well, the manufacturing side of the program may turn out to be smoother than normal.