Although BAE Systems has drawn on additive manufacturing for 20 years for rapid prototyping, using the technology for production parts only began two years ago, when the company's Tornado fighter jet flew its first metal AM part, a camera bracket. Since then, other such parts made out of ULTEM thermoplastic resin and Polyamide 12 (nylon) have been installed on flying aircraft. BAE has also used AM for non-flying repairs: protective covers for Tornado cockpit radios, support struts for working on air intake doors and protective guards for power takeoff shafts.
Machinery: What purpose did this initial use of AM parts serve, strategically speaking, within BAE?
Matt Stevens: The Tornado part wasn't a one-off. It was the start of us putting AM into more general production use for our products. It is not just for Tornado; we are looking to expand it to all of the products. In terms of non-flying parts, one example is wind tunnel models. These are functional representative parts. When we are developing new modifications or upgrades or new aircraft, they all go through wind tunnel evaluation before testing.
Machinery: What about surface finish? That has to be smooth, doesn't it?
MS: As parts come off the machine, they have a powdery finish and that does require a degree of finishing. We are investigating that. Wind tunnel models and end use applications need a smooth finish for better fatigue resistance. That is a big topic of much investigation and development. For one-off models for wind tunnels, there is not much more than hand-finishing, which is what is traditionally done anyway.
Machinery: Why are you using AM for this?
MS: For cost and time reasons. Traditionally, these small parts are machined, a job that is very fiddly and time-consuming; there is a lot of manual intervention to produce the models. With a 3D printer, you put the design into the machine, wait for it to build, do a minimum amount of finishing and it is in your hand. This is a significant reduction of development time.
Machinery: What was the story of the camera bracket on the Tornado: was it just intended to get AM started for flying duties?
MS: Yes. We have an exploitation plan for AM. In order to keep risk levels manageable, we started off with a simple like-for-like design, with a qualified material, to get flying. Having proved the process can work, the next question is how we use AM to make improved parts. That's the stage we're at now: what software is out there, how do we use the advantages that new processes can offer over conventional technologies. Then from imported part [designs], we can get on to more free-form, organic parts, where there are no straight edges.
Machinery: How soon are these designs likely to be coming through?
MS: It is hard to say a time because they are on the cutting edge. The process is developing so quickly, in terms of new technical materials and software, that it is hard to put a stake in the sand and say 'In two years we'll be doing this', because we don't know what 'this' will be in two years. Certainly I think the use of these new parts will be several years off.
Machinery: Could you tell me about some of the non-flying AM components you are making?
MS: Where plastic parts are really proving useful is at RAF bases, where some [AM] machines are deployed, for example RAF Marham, Norfolk, and not at the factory. When an aircraft comes in for a service, engineers on the base can assess the situation and build either tools or guards, or model the repair in a matter of hours to get the aircraft turned around and increase its availability.
Stevens goes on to explain that BAE's process to approve AM occurs in two stages: qualifications based on materials testing are followed by qualifications based on part manufacturing testing; only a few of each have been completed. He adds that the lack of a general industry standard for processes and materials is limiting the development of AM. "We are just moving forward with a few parts initially, until the material and process are proven to work," he says.
Machinery: Can you make any generalisations about the types of components that are most AM-friendly?
MS: AM-friendliness completely depends on the process. Metallic powder-bed parts are constrained by the size of the machines. Generally, they have a 300 by 300 mm build area, so that is certainly the biggest constraint, the size of the parts
At the moment, we are working in partnership with the industry to develop the machines and the process. We have some machines in house and we are getting some parts made at external companies and machine suppliers. During the development stage, that is quite a common process. Once the process is established, then we come to the decision on what we call 'make or buy' – whether to make parts in house or externally. We are not at that stage yet.
Machinery: What about the supply chain for AM parts. Is there one?
MS: It is quite small, particularly if we move to regular production scenarios. While this is an issue, in general the 3D printing industry is aware of this and, therefore, the wheels are turning to ensure production capacity will expand rapidly in the next few years.
Machinery: Could 3D printing really have a big effect on aerospace manufacturing?
MS: I really do think plastic 3D printing could have a significant effect, and the metal powder bed process, too. We are developing larger, free-form metallic manufacturing processes for manufacturing larger structural parts of future air vehicles, for example, where a wire feed is melted on to a baseplate, using laser or plasma. That process is limited only by the size of the machine. This can produce high quality structural parts of a free-form nature. We are looking at technology to enable this to become a reality in the future.
As an example, Stevens cites a 2013 research project led by Cranfield University in which a 1.3 m long wing spar section was built out of titanium using the Wire and Arc Additive Manufacture (WAAM) process in a single 37-hour build.
And he concludes: "If you want to talk about the longer term future, we have done some concept work looking at whether it would be possible to 3D print a complete aircraft. One of our concepts is of a mother ship flying up wherever it might be required and, depending on the threat, it can on-the-fly print [UAV] aircraft or air vehicles that will be able to be specifically tailored for the mission, in real time." ¦
