If you’re following the additive manufacturing (AM) of aerospace components, you’re probably tired of hearing about printing fuel nozzles for GE’s LEAP engine, or sensor tubes for the GE 9X—two key metal applications. And printing polymer tooling? Scott Sevcik, vice president of the aerospace business segment at Stratasys, Eden Prairie, Minn., said everyone laughed when it came up at a recent conference because it’s so commonplace they’d forgotten to even mention it earlier. But, recent advances suggest AM will continue to supplant, or at least augment, traditional manufacturing in ever more parts.
Sevcik said that although we’ve “spent the last 40 years turning metal aircraft into composite and polymer aircraft to save weight,” he still thinks the major growth opportunity to be in polymer and polymer composite printing. Sevcik and other experts also assert that high temperature, high-pressure jet and rocket engine components, and some high-stress structural components, will always be made of metal. “That’s why additive metal is so interesting for aerospace, because we’ve got no other way to take weight off it. We’ve got to keep it a metal part.”
Jeph Ruppert, director of the application innovation group at 3D Systems, Rock Hill, S.C., agreed that lightweighting components is the “slam dunk” metal AM application. He also offers examples that do more than simply cut weight. He observed that the defense and space markets would lead the charge—commercial aviation is more risk averse.
A satellite antenna bracket is a perfect case. Ruppert explained that it’s a “mass customized application.” Each bracket needs to be slightly different, because different orbits or ground stations require different antenna orientations. But, the bracket needs to be as light as possible for the 10 minutes it takes to get it into orbit, while strong enough to withstand a force of 10 G during that ascent. Once in orbit there is little or no force acting on it, so it doesn’t require long-term strength. In one such case, 3D Systems partnered with Thales Alenia Space to design and print “topologically optimized” brackets that are 25 percent lighter—yet with a better stiffness-to-weight ratio—than traditionally machined components, while making them took half the time.
Ruppert said wave guides are another important radio frequency application well suited to additive, because it’s another case in which AM’s ability to build complex internal geometries with a refined surface finish offers real advantages. Wave guides are used in communications gear in satellites, aircraft and ground stations, giving this area some growth potential. Whether it’s for electronics or an aircraft cabin air handler, “heat exchangers are everywhere,” Ruppert observed, making them another target. “If you don’t have to be beholden to the vacuum braising process [to make a heat exchanger], you can design for performance, rather than designing for the manufacturing process.”
Ruppert thinks turbo pumps and housings, “especially on the propulsion side,” will be another big application. Plus “thrusters, combustion chambers, and fuel manifolds for space.” Lest you think manufacturing for space can’t be that big a business, Sevcik pointed out that whereas previously there were only about a dozen satellite launches a year, “now we’ve got rockets launching 60-70 satellites at a time. The scale has changed considerably in space.” In fact, he said the most significant growth Stratasys has seen in metal AM has been in rocket components.
It’s harder for AM to compete with machining from solid when it comes to larger structural parts, or higher volume parts. But the crossover point is changing, said Bill Durow, manager of Sandvik Coromant’s global engineering project office for aerospace, space, and defense in Fair Lawn, N.J. One reason, Durow explained, is AM’s ability to selectively mix materials, or to print a dissimilar material onto another metal. “Take a shaft made of 300M [alloy steel], which is very typical in aerospace. If you want to add more rigidity on a knuckle where a wheel is going to be installed, you might want to add titanium. Maybe this component is going to land on an aircraft carrier and it’s going to see a lot of impact. Today you [can] print titanium onto a dissimilar piece of material to add that strength.” We’ve seen this selective printing in turbine blade repair for years (albeit with similar material), but Durow explained that he sees it used in manufacturing now as well.
Commercial airliner seat brackets are a relatively high-volume part. Owing to the stresses they experience, such as “vibrations up in the air, people moving around in the seats, and landing forces, metals are going to last much longer than a lot of the plastics,” said Durow. But despite the required part volume, AM is starting to take over with “very organic shapes, a little bit different than you’d typically see in a bracket to hold the seat in place. A lot of it’s simply for weight reduction.” Durow added that the increasing size of powder bed machines is contributing to this shift. “Ballpark sizes are about a meter by a meter by about 20 inches deep now. You can line up a bunch of these brackets and have rows and rows of machines running lights out.” Many of the metal powders are quite expensive, though. So even lights out AM isn’t necessarily more cost effective than traditional machining. But if the optimal design can’t be machined, AM becomes more attractive as its costs come down.
Sandvik Coromant recently released a new shoulder mill called the CoroMill 390 in which the entire titanium tool body is made with the additive process. This saved weight, which helps alleviate vibration, particularly on long reach applications. AM also allowed for a complete redesign of the internal coolant channels—a design that could not have been produced with traditional drilling. Durow said the optimized design ensures the “coolant actually goes to the cutting edge when you’re spinning at high speeds.”
While CoroMill 390 is now a standard product, Durow also predicted that AM will play a key role in creating customized cutting tools. “Anytime you’re dealing with low-volume orders … and I see more of that in the tooling realm now … you often have to wait for the design. Then you have to wait for manufacturing. Then you have to create programs. This all takes time, weeks in some cases. I think the additive process can help a lot. You can store ‘blanks’ in your production facility and then just add on the features you need for that tool, based on whatever the customer needs. I can see that being the future of custom tooling.”
As we move into plastic applications, let’s start with AM’s role in casting metal components. For example, in the 3D Systems QuickCast process, a shop builds a pattern with a stereolithography (SLA) machine, drains the remaining liquid when that’s complete, and then coats the pattern in a ceramic shell. Once the shell has hardened, it is fired in high heat to burn out the pattern, creating a mold for casting a metal part. With 3D printing, the shop can go from a CAD file to a pattern in a day, much faster than the process of machining a wooden or metallic pattern. Plus the process provides greater geometric flexibility.
Ruppert of 3D Systems envisions “massive growth potential because single crystal castings or high criticality castings are time consuming and low yield. There are not that many people who can do it.” Another factor, Ruppert added, is the ability to combine multiple patterns before ceramic coating to create very large molds. “You’re not constrained by the size of the machine. You can build parts of your patterns and link them up to build large casting patterns for really large parts. Since individual segments are not heavy, they’re easy to manage. So that makes that whole process enticing for a lot of different applications.”
Planetary Resources (now owned by ConsenSys) put QuickCast to work building the fuel tank for the Arkyd 100 satellite. “There is limited real estate on that satellite, and they need fuel,” explained Ruppert. “Rather than a standard fuel tank … with a pressure vessel of a welded together clamshell, they were able to print something uniquely suited to the requirements of that spacecraft. They were able to take advantage of the extra nooks and crannies… It’s a great example of the design freedom you have when using AM.”
3D polymer printing has long had a role in producing tooling for the aerospace industry. For the huge stuff, like layup tooling for large composite aerostructures, look to the Large Scale Additive Manufacturing (LSAM) machines built by Thermwood in Dale, Ind. As U.S. Account Manager Brent East outlined, the LSAM platform is 10' (3.048-m) wide and the length is extendable in 10' increments. “We’ve had inquiries for machines up to 100' (30.48-m) long, but the 40' (12.19-m) long machine is currently the longest we have sold and installed.” The machine uses fused deposition modeling (FDM), laying down a large bead 0.83" wide x 0.2" (21.08 x 5.08-mm) thick, followed by a compression wheel that fuses the layers together and helps ensure precise print bead dimensions. Because the method does deliver a ribbed surface, the LSAM also features a router for final machining to the desired finish.
East added that the size of the parts that can be printed, and the speed at which they can be printed, aren’t limited by the LSAM print head, which deposits over 500 lb (226.8 kg) of material per hour. It is limited by the need to apply each layer when the previous layer is at just the right temperature to ensure proper bonding. “We monitor that with a thermal imaging camera,” he explained. “And the whole process can be controlled directly at the machine, which is key. Because, if you’re in the middle of running a part and not entirely happy with the result, you can adjust it on the fly, versus other software that forces you to go back to the office, figure out where you were in the program, change your settings, repost that code, and then try to get back to the machine before your part is scrapped.”
LSAM machines print a variety of materials, depending on the application. “Most of the room-temperature trim tools and those types of fixtures or holding jigs are printed with ABS pellets,” recounted East. “We’ve run pelletized fiber-reinforced plastic, polycarbonate, and nylon. We’ve also used PESU (polyethersulfone resin) and ULTEM.”
Thermwood has introduced an optional vertical table that enables continuous printing across the entire length of the machine. In this configuration, the nozzle is turned 90° (so it’s horizontal), and the servo-controlled vertical table indexes away by the thickness of the bead with each layer. East said this process is slower than printing smaller segments horizontally and then splicing them together, which is fine for room temperature parts. But Thermwood found that such segmented parts tend to delaminate at the joints in an autoclave because the adhesive breaks down. So for those applications, vertical printing is the solution. East added that geometric considerations might also dictate vertical printing. “Each polymer has a different coefficient of thermal expansion, and the expansion also varies in X, Y, and Z; it’s anisotropic. Depending on the particular end use of the part, you may want the largest amount of growth in a certain direction, which could ultimately dictate printing horizontally or vertically.”
Sevcik from Stratasys called air ducts “one of the sweet spot applications for laser sintered nylon printing.” Air ducts are numerous, complex, and varied, each with slightly different bends and radii. But small ducts have been well addressed, so he doesn’t think there’s much growth potential. On the other hand, he offered that FDM printing has driven an “immense amount of growth on larger components.”
He pointed to projects that combined multiple builds of 3 x 2 x 3' (0.91 x 0.61 x 0.91 m) to create large duct components. For example, “United Launch Alliance makes a ring of ducts around their Atlas V rocket fairing to blow cold air onto electronics on the launchpad to keep it cool prior to launch. It is a 5-m ring produced with 16 of these ducts, 3' (0.91 m) at a time.” The prior approach required assembling 144 aluminum components. “The Air Force has a mixing chamber, like a plenum, where ducts come together, that requires five builds to produce a very large part for the C5 aircraft.”
Both Sevcik and Ruppert see electrical connectors as another growth opportunity, especially with digital light processing (DLP) technology. As Sevcik put it, “DLP is the opposite of FDM, allowing us to go smaller, with finer resolution, for things like connector components. Connectors are incredibly variable, just like ducts. Being able to go down to a scale with the right materials gives us a whole new application space to grow within the aircraft.”
The experts agree that newer materials will expand 3D printing. These new materials include ULTEM, carbon fiber reinforced PEKK (polyetherketoneketone), and Antero (a PEKK-based material). It is conceivable they may replace metals for some applications. If so, we’ll get what Sevcik calls “a double whammy: We get AM’s topology optimization to reduce the weight, plus we get a lower-density material.” It’s also conceivable that 3D metal printing might do more of the “must-be-metal” components, like turbine blades and landing gear struts. While that wouldn’t cut as much weight as replacing with a polymer, it would still offer improvements.
The key, according to Ruppert, is achieving a “satisfactory level of reliability and repeatability… As things like machine learning and in situ monitoring improve, for all AM, we’ll see an increase in the ability to work within a process window that’s been well characterized, rather than just parameters. That will allow us to also work with materials that may change their printability as the different types of parts you print on it grow. Being able to make on-the-fly changes in parameters based on your characterized process window because your thermal differential for layer X is different than it was at layer Y.”
In other words, if a machine shop can fully monitor and control its build process and make it repeatable, it can certify the process. And if it has certified that the parts meet the needed standards, it can use its newly certified AM to produce them. Ruppert thinks we’re moving in this direction, though he added that such a process is probably more than five years off.
Keep your subscription to ME!
Connect With Us