Technology and lessons learned from the A380, such as the assembly of the composite vertical tailplane (VTP), were plowed back into the A320 as part of an overall improvement which culminated with the development of a new single-aisle, lean manufacturing-based moving line at Hamburg in 2006. With the rate expected for the single-aisle line it was “not possible to stick with dock manufacturing in the assembly process. So we changed the production philosophy for the fuselage and wing from a stationary process to a moving line. Based on lean manufacturing principles, this gives us better visibility on where we are with parts, the supply chain and deliverables. With an increase in production, the supply chain increases too and missing parts become more critical,” says Gralfs, who notes that an additional challenge was “to make that transfer in the midst of production,” a process that will be repeated to a lesser extent with the phasing in of the NEO from 2015 onward.
The single-aisle line in Hamburg produces one fully assembled fuselage every 7 hr. Compared with the rates on the original line, the implementation of lean processes has cut lead time by 33%, while at the same time producing a 50% rate increase. “We are also doing the same at Saint Nazaire, [France], (for the forward fuselage) and for the wing line in Broughton, [U.K.],” he adds.
To help prepare for the A350, now entering production in Toulouse, Airbus has evolved the “digital factory” concept from early steps taken with the A380. As a result, a wide-ranging set of new techniques and a greater level of automation has been introduced, says Gralfs. The digital factory integrates design and manufacturing in the industrialization phase, reducing costs from concept to implementation. The routing of parts around the production flow, specification of the jigs and tools and details of the process structure were all simulated and verified in a cyberworld prior to execution in the real world. “We used [simulation] a lot in the A350, which was a dramatic change from 10 years ago when we were developing the A380.”
“We start with key input from the digital mock-up (DMU) to perform virtual process planning. We need to do it because the ergonomic aspects are key. That enables a faster ramp-up, an increase in human work efficiency and lower health costs. Staff are aging and people are working for longer [periods], so accident prevention is important. The A350 wing is extremely challenging from an ergonomic perspective. The stringers are more like rails on a railway so we need to automate that and find solutions to allow people to gain access to difficult areas.”
The “master” DMU brought together all the partners working on both design and manufacturing. Throughout the design process, the DMU was constantly being updated to ensure configuration control. Unlike the A380 program, in which assembly delays occurred because multiple DMUs and design tools were used, the A350 has relied on a standardized set of software tools across everything from production, engineering and finance to composites design and other structural aspects. The single A350 DMU continues to be maintained as the program moves into serial production.
Other design features of the A350 drove the development of sophisticated methods of precise positioning and alignment. The fuselage of the A350, for example, comprises large composite panels attached to frames and stringers, rather than just one-piece barrel segments, as with Boeing 787s. “If you have composite sides you can run into a nightmare in assembly because they have a tendency to not be as accurate as metal parts. With the A350, the challenge is mounting those shells with no shimming, which demands high accuracy. So how do we do it without hundreds of tons of steel tools? We set up an NC [numerically controlled] measurement-assisted assembly system based on DMU to bring it together. We are not using fixed frames and, instead, have NC axes giving us positional accuracy for handing the shells,” says Gralfs.
While the fuselage assembly methods differ between the A350 and 787, the two competitors have adopted similar methods of pre-equipping large modules to reduce lead time and recurring costs. An example of this is the decision to pre-install systems such as electrical harnesses and other cables and piping above the avionics bay using a secondary support grid structure, rather than attaching them to crossbeams with fixed brackets in the traditional manner. “Lead time has been reduced by 80% on the floor grid in the nose above the avionics bay. Before, it was impossible to produce as a module so we started to go into the architecture. As a result, we developed a secondary structure we can pre-equip,” says Gralfs, who adds that the integration of the unit lies on the critical path.
Increased use of automation is also expected as Airbus drives up delivery rates and improves quality. Among the manufacturer's recently automated production tasks is the process to make VTP reinforcement fittings at the base of the fin. The procedure was formerly done by hand and involved draping, pre-forming, handling and cutting. Now the automated resin transfer molding process is making 5,500 fittings per year with improved quality and a 50% reduction in lead times.
Another automated system is a small robot, developed by Spanish manufacturing equipment specialist MTorres, which “walks” over the fuselage, holding itself in place by means of suction cups on its feet. The unit, called the Flexible Drilling Head, is a crawling-drilling-riveting machine that locks itself in position to drill and rivet before releasing itself to move at a speed of 3.5 mm per min. to the next position. The 5-axis crawler, which weighs 220 lb., can drill and rivet circumferential, longitudinal and conical joints. In use on the A380, the robot drills 8,500 holes per aircraft as it moves around Section 19 of the fuselage. The system has helped generate a 45% cut in lead time, and is capable of drilling into titanium and Glare, a lightweight commercial glass laminate aluminum-reinforced-grade material, as well as composites.