The devil is in the details: So, you 3D-printed a part for a jet engine part. Now what?

3D printing has rightfully gotten a lot of buzz because of the marvels it can do. Also known as additive manufacturing, it has opened new paths for designers to create custom shapes that were previously too expensive or downright impossible to make. The technology’s potential is enormous, but GE engineer Peter Martinello offers a dose of perspective. “This is true if you have to print just one part,” he says. “But as soon as you have to print thousands of copies of the same part consistently, you are in a territory that’s both similar to traditional manufacturing, but also more complex than anything that we’ve had before.”

Martinello knows this as well as anyone. As a senior engineer for additive design at GE Aviation’s Additive Technology Center (ATC) in Cincinnati, he helped bring to mass production in 2016 a 3D-printed fuel nozzle that efficiently sprays fuel inside the LEAP jet engine. This wickedly complex metal part the size of a walnut was previously made from 20 components. GE now prints 600 of them per week as a single piece.

But Martinello and his colleagues have just taken that feat to a new level. Over the last 10 months, they have developed a new process to mass print a much larger part for a version the GEnx jet engine that powers the latest generation of Boeing 747s. To do so, the team also had to “production-proof” the new line of printers for the part.

This is the first time that GE has designed a mass-production process for a line of its own printers made by Concept Laser, a German engineering company it acquired in 2016. “Minor tweaks here and there are OK in the development phase, but when you get into production, everything has to be locked down,” says Danny Brandel, a lead manufacturing engineer at the ATC and a member of the team. He says this is especially important in Aviation because of the tight regulations of the FAA due to the importance of safety and quality in air travel.

GE Aviation plans to begin mass producing the GEnx parts this summer at its factory in Auburn, Alabama. “The reason we did this project was because it represented several firsts for us,” says Eric Gatlin, a general manager for GE Aviation’s additive integrated product team. “It’s our first program we certified on a Concept Laser machine, and it’s also the first project we’ve taken from design to production in less than 10 months.”

Top: Brandel (left) and Martinello (right) helped to get ready GEnx engine brackets in their hands for 3D printing on Concept Laser machines. Above: A device for sifting metal powder at the Additive Technology Center in Cincinnati. Images credit: Tomas Kellner for GE Reports.

But it almost didn’t happen. Designers typically use 3D printing to make parts with complex shapes like the LEAP fuel nozzle or sections of the Catalyst turboprop engine. Since they can print pretty much any shape they desire, the technology allows them to combine multiple parts into one. The Catalyst, for example, which was also partially developed at the ATC, coalesced 855 parts into just 12. The result are components easier to maintain because there are fewer parts to inspect. The technology can also radically shrink the supply chain, since all the factory needs to worry about is a steady supply of powder for printing, rather than having the right size of nuts and bolts in stock.

But the part for the GEnx, a simple bracket the size of a human rib that holds open the engine cover during servicing, had few of these qualities. It already was a single piece of metal, and because it’s already inside a working engine, the GEnx first flew in 2010, the designers had to stay close to drawings approved by the FAA.

That created some hand-wringing. First, members of the Additive Global Management team, who decide what print, had to agree whether they even want to make the part. Every two weeks, engineers pitch their ideas to this diverse body, which includes employees from GE Aviation’s supply chain, engineering, and engine product lines. “When the project first came to us, we said ‘this is an oddball, this is an outlier,” Gatlin says. “It took time for it to sink in, before we realized that this is exactly what additive can be used for, to demonstrate its speed and low cost.”

The first GEnx engines used brackets made with traditional methods such as milling. “When the GEnx program kicked off, they just hogged the brackets out of a big block of metal,” Martinello says. “By the time you had the finished product, you cut away more than half of it.” The team realized that by 3D printing the part, they would we able to reduce as much as 90 percent of the waste. “We just had to machine some bolt holes and some clevis pins and we were done,” he says.

The new approach would not only allow GE to produce the part in-house, but also to reduce supplier expenses. “We were being charged a significant amount for these parts,” Gatlin says. The team also implemented small design tweaks that reduced bracket’s weight by 10 percent. When it comes to flying, every ounce counts, he says.

A GEnx engine on a test stand in Peebles, Ohio. Image credit: GE Aviation.

But the project was only half done. Working in parallel track, a team of engineers was readying the Concept Laser machine for mass production. The group, which was based in the U.S. and also in Germany, would get together every day on a 7:30 am conference call to track their progress and lay out their next steps. “It’s a highly collaborative process,” Martinello says.

They selected Concept Laser’s M2 printer, a fast, mid-size machine that uses a pair of lasers to print four brackets at a time. One of the team’s tasks early on was making sure that brackets coming out of the machines were of identical quality. For example, they took a system designed to air out the smoke produced when the powerful lasers hit and melt the layers of powder into the desired shape, and gave it more power. If left untreated, smoke could deposit soot into the tiny pools of molten metal created by the lasers and alter the density of the part, but also disperse the laser beam and make it less sharp, like sun rays filtering through morning fog. “Air flow is pretty critical in terms of getting parts with good internal and external quality,” Brandel says.

Brandel and his colleagues spent several months making sure the machines were ready for prime time. “I like to get my hands dirty and get to understand the inner working of the machine, rather than just churning out hardware,” he says.

To keep an eye on the conditions inside the machines, the engineers installed sensors and high-resolution cameras monitoring the power and the stability of the laser, the oxygen levels in the printing chamber and other factors. They started by focusing on individual machines, but in the future, they will feed the data into the cloud, monitor the parameters across a whole fleet of printers and look for patterns that will allow them to spot potential problems early. “We want to have everything quantified and take variation out of the process,” Martinello says.

Getting the part ready for printing was another riddle. To move faster, the team decided to print the bracket from a cobalt-chrome alloy, rather than the original nickel-based superalloy called Inconel 625, because the FAA already certified it for use inside engines. But in order to make approach economical, they had to print four brackets at a time. That’s when a group designers working at the ATC came up with a clever solution: fitting all of the parts like interlocking fingers onto a single “build plate” the size of an average computer screen. “It looks like one of those wood block puzzles, where all the pieces fit together,” Gatlin says. “When you are printing it, it’s hard to tell whether it’s one part or four parts, but when you cut they off the plate, they separate and you have an aircraft’s worth of brackets in one build.”

The engineers from ATC and Auburn spent the spring collaborating and comparing notes so that Auburn will be ready to start cranking out the brackets. The list includes printing parameters, but also heat treatment and inspection steps. “There’s no part that just comes off a printer and goes straight into the engine,” Gatlin says.

Back in Cincinnati, Gatlin, Martinello, Brandel and their colleagues are already working on their next projects. They’ve identified more than 100 components that could be printed. Of those, one-third are new products like the Catalyst engine, but the rest involve redesigning existing parts like the bracket — a huge new market for additive manufacturing. For example, they are working on significant cost and weight reduction projects across their other engine lines like the LEAP, the GE9X engine for Boeing’s new 777X widebody jet, and military programs.

3D printing is still so new that Martinello, who is 37, is often learning on the job. “I’ve been doing this for three years now, but I still have to take a step back and remind myself what it is we are trying to do and why,” he says. “Fundamental engineering principles still apply. It’s just a new tool that you have at your disposal. It’s about understanding what it can do, and how it can improve your designs.”

This story was originally published on GE Reports.