ATLANTA--Collins Aerospace is taking the next step with its aerospace additive manufacturing activity, with aspirations to develop 20 production-ready metal parts in 2019, says Paula Hay, the company's executive director of additive design and manufacturing.
"We think we'll have quite a few parts flying this year," she told MRO-Network in an interview at Aviation Week's MRO Americas. "It's a high challenge, when you consider that before this year started, we had one in production."
The company's first complete part made via additive manufacturing (AM) is a fuel nozzle for an industrial gas turbine. Its second, now being delivered to a European military customer, is an aircraft part. Developing both helped Collins learn many lessons, but one major step—certification by a civil aviation regulator—was not necessary.
While some suppliers have AM parts flying, FAA and other regulators have not developed specific guidance for certifying them. AM parts have to meet the same regulatory standards as any other part, but demonstrating how AM parts meet those standards is a challenge.
"You certify the part exactly the same way," Hay says. "The question is, how much more information do you need to convince people that the part is a good part? It still has to meet all the same requirements, but there is more information you need to make people comfortable."
Another issue that AM parts introduce: the need to ensure each part is the same, no matter where it is made. One of the great attractions of AM is the concept of buying a machine, acquiring a part's design and data, and producing parts wherever its needed—such as in an MRO hangar. But right now, regulators are reluctant to consider two identical machines as part of the same production process. That means each machine needs its own approval, even if its producing the same part as another machine. The issue gets more complex when different brands of the same machine are introduced.
"Right now, regulators are favoring serial number-based approvals," meaning each machine must be certified as part of a production process," Hay says. "We're getting closer" to gaining approvals that cover multiple machines built to the same specifications, she added. "I think we'll get there before we get machine-agnostics tied to two different brands."
Working through such challenges doesn't happen in a research-and-development (R&D) environment, Hay says. That's one reason why Collins has spent the last year focusing on making parts for potential customers, not exhibit-hall display cases.
"If you keep doing the R&D work, you never really tackle the big problems," she says.
Collins is not delving into safety-critical parts—at least not yet. Among the areas Hay sees as ripe for AM parts: heat exchangers, housings, valves, and certain brackets.
The company's customers, primarily the aircraft manufactures, still need convincing, but they are coming around, Hay said.
The parts chosen for AM will not simply be "drop-ins" that replace a traditional parts just for the sake of changing things up. "The real value comes when you bring the engineering side into it," she says. Targeting parts with reliability issues or opportunities to save weight are high on the list. In some case, parts can simply be designed better when AM is the production method. That requires getting buy-in from engineers that have done things a certain way for a long time.
"That duct doesn't have to have right angles. It can be curved if you want it to be," Hay says, offering a theoretical example. "They have to learn to think differently."