When Airbus Hamburg looked to establish additional rate capacity for A320 final assembly in a 4th line, a number of factors dictated that a direct replication of the existing FAL capability seen at the four global A320 final assembly sites in Toulouse, Hamburg, Tianjin and Mobile was not an option.
Frankenberger was challenged with creating additional capacity for A320 final assembly against two significant constraints: it had to happen fast and it had to fit into a hangar which had yet to be vacated by reduced A380 production. A structural partition wall meant only narrow access between the two halves of the hangar, just large enough to pass a completed fuselage through.
The standard A320 FAL is essentially unchanged from the mid-80s, using large hard iron jigs, necessitating heavy foundations never designed for the current, increased production rates. This was a chance to address a significant reduction in non-value adding activity, specifically in the moving and positioning of parts.
Production equipment and infrastructure suppliers met for the first time at the Paris Airshow in June 2015. In April 18, the first aircraft came of this new assembly line and by October 18, rate 5 had already been achieved. Even without any significant changes, this is a reasonably fast capability and equipment procurement and ramp-up cycle. The line is now continuing its ramp up and is designed to be rate 10 capable.
The major fuselage joint between the forward and rear sections is achieved by ‘Luise’ and ‘Renate’, the affectionately named left and right joining robots. These are well-proven Loxin automatic drill and fastening heads mounted on heavy Fanuc robot positioners, however the jig reference point is a laser beam which shoots straight down the fuselage centreline.
The heart of the matter
The heart of the 4th line is what Frankenberger refers to as dynamic measurement assembly. Metrology is the factor which makes this work. The single biggest innovation is the change in referencing system, from alignment against hard tooling to one based on accurate measurement. This allows a transfer of the data reference from the floor plate to a more directly relevant one on the outside of the aerodynamic surfaces. This relies on a cascade of three orders of magnitude for metrology and movement accuracy, within 2cm for the battery powered mobile tooling platforms (MTP) moving major components around the hangar, 2mm for the MTP’s approach to a referencing line, and 200µm in aligning the airframe sections to each other. The MTPs becomes part of the assembly station because they are sharing the same geometrical reference system.
The hangar floor is a standard industrial one, with undulations of a few centimetres which would otherwise cause problems. Whilst the MTPs are in motion, though, the ride is described like being “on a flying carpet”. There are load cells everywhere which are carefully monitored and controlled to avoid introducing stresses into the incomplete airframe. This is entirely consistent with the whole philosophy within the line, though, with an ability to accommodate geometric uncertainty rather than to fight against it.
There was an element of nervousness surrounding the stage where the MTP wing supports are pulled away and the incomplete airframe is transported a few metres to the next station supported only on its fuselage. The synchronisation of MTP movements is carefully choreographed to avoid putting excessive stresses onto the fuselage which would cause unacceptable deformation.
The principle of operation is to deliver all materials straight to the point of use in a 3m radius of the operator enabling workers to conduct the tasks, and to have their tools, consumables and the aircraft parts in direct access. Frankenberger likens this to the work environment of a surgeon or dentist and predicts that some ergonomic aspects which have been tested here could subsequently be transferred to the legacy lines.
The reduction in non-value adding actions is evident, and significant, with much of this drawn from getting positioning right first time and reducing rework and the consequent manual adjustment which goes to making parts and assemblies conform to rigid fixtures. The best fit alignment of major assemblies through measurement and automatic actuation is a major improvement on the legacy trial and error, or test and adjust methodology and the whole line has an uncluttered nature.
The measure of learning
Frankenberger is aware that one size does not necessary fit all, though. The engine pylon is offered up to its alignment pintle and mounting brackets on the wing using a series of actuators on its MTP, but in reality, this job would be more efficiently achieved manually by the use of a local crane or suspension point.
Automation has provided a rich vein of manufacturing data. Frankenberger confirms that “when you start measuring, you start really learning”. The real production data which has been building at rate 5 since October last year is evidencing hotspots of tolerancing factors which provides an indication for upstream manufacturing and major assembly processes.
“Many of the processes here are having implications upstream. As soon as you have credible data in how the major parts are coming together, this will have an impact on how these are made further up the build sequence. There is also the potential to learn how to design to be automation friendly.”
This has given Airbus, and Frankenberger specifically, an appetite for greater assembly automation. There a ‘step change’ project in Hamburg looking into the single aisle system where innovation is considered possible in terms of optimising the handling positioning the fuselage section build.
He expands that even with robots, including the MTPs, providing the positioning accuracy, there remains a need to allow human access to conduct the joining processes and so there is still a need for all the added complexity of stairs and platforms and staging. However, thinking beyond the current case and the ability to design an aircraft for automated assembly processes then there would definitely be opportunity for more flexibility, in the flow and in the different sizes, going to a multi-programme line if necessary and also being flexible to move a line to another hangar or change within a hangar. This all comes from having the confidence to let go of the rigid geometrical reference points.
This line could feasibly be driven, mostly under its own motive power, into any hangar on the Hamburg site. The ability to replicate and deploy this assembly line, with a relatively modest investment, enables a step change in the scalability of this and similar manufacturing operations. There is little in the fundamentals of the dynamic measurement assembly methodology proven here which could not be deployed up and down scale in both the size of aircraft products and in the production rate requirement.
Another lesson for the future is not to overload each station with different functions and production steps and to keep each step simple and uncluttered. The conventional way is to have a positioning function and also to have the final joining within this, which are two very different manufacturing operations with many aspects of conflicting requirements. Frankenberger postulates that the equipment or capability redundancy, for example in the use of the metrology system which provides the positional accuracy and is then tied up whilst the joining operations are carried out, can be reduced by designing the whole line around this new metrology setup. The time for automated positioning is about 5-10% of the lead-time of the station. The ability to freeze or fix the positioning and then to have more unconstrained joining automation within another station is foreseen.
The digital revolution
The 4th line is part of an overall Airbus agenda of revolution in digitalisation in manufacturing. Digital Design, Manufacturing and Services (DDMS) is a company-wide transformation project enabling state-of-the-art digitalisation and the seamless flow and use of data to provide a major efficiency breakthrough across the programme and product lifecycle. Frankenberger talks about the principle of digital continuity, which is highly evident within the 4th line.
Once digital capability is expended outward, into a larger ecosystem, it starts to influence wider business decisions and Frankenberger admits this is making his life as an industrial architect much more interesting. The product, industrial system and operation of the aircraft have to be conceived as a single data entity, to match with the business objectives for aspects like lead-time, flexibility, and customisability. DDMS is basically a holistic approach where there is a system of systems, bringing in life cycle considerations.
The fourth line has been a culmination and a coming together of quite a few different thoughts. The tail wind to the realisation of the 4th line was the ramp up in rate and the constraints of an existing hangar space. Without the constraints imposed on this, the innovations would not have happened, but then necessity has always been the mother of invention.