BAE Systems: behind the scenes of the F-35 Fighter

Posted on 23 May 2017 by The Manufacturer

The Manufacturer was granted a rare look behind the scenes at BAE Systems’ Military Aviation and Information (MAI) division at Samlesbury in Lancashire, to witness the production lines for the F-35 Joint Strike Fighter and the Eurofighter Typhoon. Nick Peters reports.

The f-35 Joint Strike Fighter (JSF) is the world's largest defence programme - image courtesy of BAE Systems.
The f-35 Joint Strike Fighter (JSF) is the world’s largest defence programme – image courtesy of BAE Systems.

The BAE Systems plant at Samlesbury is home to more aeronautical ghosts than just about anywhere else in the country. Wander around the old airfield, once destined to be the airport for Blackburn and Preston, and today you see the massive factory units where today’s manufacturing takes place.

Listen back through the years and you’ll hear them making Hampden and Halifax bombers during WW2 (the women technicians turning out three aircraft a day.) Post-war came the English Electric Canberra bomber and the iconic Lightning fighter, a model of which greets visitors at the main gate. Given the US Air Force also deployed their own Lightning fighter in the past, the F-35 is known in both countries as Lightning II. In more recent times, Samlesbury was home to the Jaguar and Tornado.

Today, Samlesbury is a showcase for the progress of manufacturing technology over the past 45 years. The Hawk trainer (first flight 1974, most famously used by the Red Arrows to this day) is fabricated, a traditional metal-forming product. The Eurofighter Typhoon (first flight 1994) is a step-change forwards, being constructed of aluminium, titanium and carbon fibre composites. The F-35 Lightning II, the US-led Joint Strike Fighter, is built on much the same principles, but with ultra-fine tolerances made possible by modern superfine machining and advanced robotics.

It is a fascinating display of industrial might supporting military might, underpinned by a thick layer of political wrangling and cost overruns. Much of that wrangling is ancient history, but it has modern echoes in President Donald Trump demanding that the F-35 JSF programme’s costs be slashed. (For the record, Lockheed Martin, the F-35 JSF’s lead partner, says cost-cutting was already underway, but with political astuteness credits the new president for encouraging and hastening the process.)

The issue of getting more bang (if you’ll pardon the expression) for fewer bucks is not unique to the American president. The UK government is continually shaving, some might say slashing, the defence budget, so BAE Systems is under pressure from both national governments. The problem, however, is how to introduce cost cutting into production processes that were developed when the programmes were first launched.

As far as the Hawk is concerned, output is now limited to about a dozen planes a year to export clients, so the need for urgent evolutionary change is not so pronounced, although BAE engineers are always alive to ways in which the limited run can be enhanced and improved by new technologies.

It is the Typhoon and F-35, very much live MOD acquisitions, that offer the greatest opportunity – and imperative – for replacing parts that combine the daunting mix of intensive man-hour manufacturing with low lifetimes, with parts that can only be made possible by technologies that were invented long after either plane was a mere twinkle in its designer’s eye.

This article offers some illustrations of how BAE Systems has successfully deployed these new technologies in the search for greater efficiency, lower costs and longer operational lifespans (and a happier President Trump).

The F-35 Lightning II machining facility - image courtesy of BAE Systems.
The F-35 Lightning II machining facility – image courtesy of BAE Systems.


F-35 production

The last thing you notice about the main F-35 production sheds is the ambient temperature, an unremarkable 21ºC. It is actually mandated by the plans, and if you were to visit the other F-35 production centres around the world, in Texas, Italy or Japan, there too it is 21ºC.

The point is to allow zero opportunity for the different expansion rates of the materials that go into the aircraft, mainly aluminium, titanium and composites, to affect the end product.

What is immediately apparent, however, is that measurements are in US Customary Units, more or less equivalent to our Imperial System. So one walks from the metric world of the Typhoon sheds into a world of feet and inches in F-35. (This means that students at BAE’s Academy for Skills and Knowledge, also at Samlesbury, where they train the aeronautical engineers of the future, must be fluent in both systems.)

BAE Systems makes the rear fuselage of the plane, from just before the tail backwards, to include the two V-angled vertical tail-planes and the two horizontal tail-planes. There are 3,180 F-35s on order, so there is plenty of high-value work to be done in the coming years. The contract extends to a range of internal systems and post-production services.

There is something deeply impressive and simultaneously quite intimidating about the main machining facility. Sixteen Starrag 1250STC 5-axis milling machines dominate the space. Behind glass windows one can catch glimpses of the robotic arms sliding between the machine head and the tool storage to select the right one for the next phase of the job.

The titanium billets come in pre-cast by subcontractors, in rough form, ready for the fine machining that takes several hours per unit. No room is allowed for error, with tolerances of just 5/1000ths of an inch in the machining process, which drops to just 7/10,000ths of an inch during the assembly process. It is a financial as well as an engineering imperative – one intricate part of roughly 3 feet by 2.5 feet is dubbed the DB9, because one could buy an Aston Martin for the same money it costs to produce.

Just as with the Typhoon machining, human intervention is minimal. A small handful are on shift at any one time, 24 hours a day (the 1250 STCs have a 95% average uptime), their tasks limited to monitoring the machines, loading the raw product into the Starrags and removing the finished parts, which then go to the one unit that does still require human interaction, de-burring. This just ensures that the edges of the parts are completely clear and sharp. They are then placed into a Heckert HEC 1000 jig borer, also made by Starrag, for the holes to be drilled for the eventual fitting of the composite skin.

From machining they then move to the assembly facility, a super-clean FOD-free environment. Cradles containing the developing tailplane sections move by pulsing along the line (currently once every three days, but will reach once a day at peak rate), giant cranes lifting and carrying them from bay to bay. All the tools required for each process are in place, and the cradles are rotated to allow ease of access. Close inspection of labels allow one to identify which particular plane is being made at any one point. The one I followed was destined for the South Korean air force.

It takes 42 days to complete each fuselage/tail-plane structure, from the machined elements entering the line through to assembly, composite addition, hydraulics and electrics, painting and testing. It ends up being placed on a secure trestle for air freight to the final assembly lines in the US, Italy or Japan.

Airframes and composite skin

The world of aircraft construction changed when technology made it possible to move from whole-metal fabrication into metal airframes surrounded by carbon fibre composite skin. Both the Typhoon and the F-35 share the same process, but the relative newness of the F-35 means it’s less time and materials-consuming.

Key to all modern aircraft manufacture is ICY – interchangeability. Any part must be capable of being swapped out for a replacement at any time, which calls for total accuracy. The fact that different parts of the aircraft are being made in different plants across the world underscores that need.

On the Typhoon, the aluminium airframe elements come out of the Ecospeed F-fitted 5-axis machining centre, after finishing that can take up to eight hours for each side of the central fuselage frame. The metal, which constitutes just 15% of the airframe, is laser-analysed, as is the composite skin to which it will be attached, to measure what gaps need to be filled to achieve ultimate accuracy. The frame then receives a coating of Hysol epoxy resin, which is machined to create the exact shape required. Holes are drilled in the frame and the composite skin sections are fitted and further milled to reach 100% ICY conformity.

The process is the same on the F-35, except the tolerances to which the frame has been worked are so fine that they enter the production line with the fitting holes already drilled. Instead of using the Hysol process, a thin layer of green glass-fibre coats the side of the metal that will receive the skin and this is what is milled to achieve ICY. In both aircraft, a high-grade petro-rubber chemical sealant is used to make the join.

Working at 900ºC on the superlastic forming press - image courtesy of BAE Systems.
Working at 900ºC on the superlastic forming press – image courtesy of BAE Systems.


SuperPlastic Forming and Diffusion Bonding (SPFDB)

As the name implies, this is about treating titanium as if it were plastic, heating it to the point where it can be shaped and formed precisely to design. The precise temperature, I am told with security-inspired vagueness, is ‘close to 900ºC’.

The process is used to create the foreplanes that sit on the nose of the Typhoon, which help deliver its incredible mid-air agility. The aircraft’s avionics systems move the foreplanes anything up to 50 times a second, which requires both lightness and strength. Hence the use of titanium.

The foreplane starts life as a flat sheet of titanium which is milled via a 3-axis machining process to the correct thickness and then cut to produce four sheets that form a ‘pack’. Between the two middle sheets, a screen printer (‘Just like in the T-shirt shop!’ one of my guide’s joked, although he’s not far from the truth – the difference is only one of precision) spreads a compound of yttrium oxide to ‘stop off’ the parts that are not intended to stick together when the bonding begins.

The pack is then inserted into a press where the combination of c.900ºC heat and 300psi pressure creates, over a period of many hours, a bond between the sheets, except in those areas that had been printed, or stopped off. Diffusion bonding is not about melting the metal, but creating the condition for its atoms to interact and bond in a way that makes the product indistinguishable from original metal to all but fine metallurgical analysis.

Gas nozzles are welded into the pack, which is then transferred to a mould in the SPF press, and as the metal again reaches its c.900ºC temperature, argon gas is very slowly injected via the nozzles. The pack expands to fill the mould over a period of several hours. It’s cooled, then rigorously scanned and X-rayed to ensure conformity with the design, before final cleaning and machining.

In the early days, failure rates on the foreplanes were high, which at a cost of £250k was a serious problem. The process of reversing this, so that there is nowadays a 96% success rate, is down to a mixture of improved technology and an acceptance that the mandated tolerances in the early days were far too stringent. They were based on a ‘defect free’ mantra, which experience has demonstrated is unnecessary.

It also means that the lifespan of products like the foreplane that have been manufactured in this way has jumped from 600 use-hours to 6,000 use-hours, which saves considerable money over the life of the aircraft. The process is also very light on labour. There are five giant presses in the SPFDB shed, requiring just four operators on any one shift.

laser melting (SLM). Currently, this part comprises several separate pieces of metal - image courtesy of BAE Systems.
A prototype replacement air intake duct for the Hawk, printed via SLA, with a view to being produced in titanium via selective laser melting (SLM). Currently, this part comprises several separate pieces of metal – image courtesy of BAE Systems.


New Product & Process Development Centre (NP&PDC)

The vehicle for adapting legacy manufacturing systems at BAE Systems and injecting enhanced processes is the NP&PDC. Its manager, John Dunstan, is a patient man and he needs to be. The imperative to change, from the viewpoint of any shop floor for whom change is usually a pain, is limited, and constrained by original product qualification, so Dunstan and his team have to act as ambassadors and advocates for the processes, and the benefits they can offer. “We were a process looking for problems!”.

BAE was quick to join the 3D printing revolution and makes a significant number of parts each year in this way – 2,500 in 2016. But Dunstan and his team are seeking much wider understanding among their colleagues of the capabilities of the new additive technologies, particularly metal powder. Dunstan showed me a spindly-looking metal device that had been printed as a potential replacement for a bracket to fit on the Tornado. He had been able to prove to the designers that the new bracket was not only much lighter than the one it replaced, but didn’t require fabrication. It just quietly printed itself overnight.

The team had similar success with a cooling duct for the Typhoon that was not only lighter, and made of one piece of plastic rather than several different rota moulded glass cloth pieces, but was significantly more efficient because they had been able to introduce a design change by taking full advantage of CAD modelling features that the legacy process could not accommodate.

Even something as simple as a map box (yes, they still use them) by the side of the Hawk pilot’s seat used to take up valuable man-and-machine time, but can now be printed in plastic and slotted home.

“This comes into its own when you have several fabricated and welded parts replaced by one part of often better design and which has ironed out original issues and problems,” John Dunstan said.

The future importance of Dunstan’s department will become more apparent as aircraft output volumes decline and specialism increases. Simply put, it is not economic for many component manufacturers to bid for the work when the civil aerospace sector offers so much more volume.

“We can build features into components that could never have been machined,” Dunstan added. “It means we will in the first instance take traditionally manufactured items out of the supply chain and bring them in-house, to establish process capability.”