In additive manufacturing for space hardware, the parts easiest to print are often the hardest to certify. A forging and a weld seam unified into one printed piece sounds like pure upside, until someone has to prove it will not fail under pressure in orbit.
That was the case Andrew Thompson, Manager of Northrop Grumman’s Additive Manufacturing CoE, made at our AMA: Aerospace, Space & Defense 2025 conference. His company prints hundreds of thousands of tooling parts a year, but the talk was about a harder category, “flyaway products,” meaning hardware that actually flies on the spacecraft.
“I want to scale up this technology from brackets that save programs $5,000 and start making tanks that save programs $500,000, if not more, depending on the size,” he said.
Our AMA: Aerospace, Space & Defense 2026 online conference returns on July 9th. Register for free today!
The Inspection Bottleneck of Single-Piece Parts
Northrop’s propellant tank is the clearest example of that tradeoff between easy to print and hard to certify. “We went into this thinking, how do we make tanks as a single piece? How do we help address a supply chain challenge with both forgings and complex welding?” Thompson said.
The team built it using directed energy deposition (DED) in titanium Ti-64. The design drew on the GAMAT program’s dataset, a materials database effort Thompson credited primarily to America Makes and Boeing.
What started as a demonstration performed well enough that Northrop moved it into formal performance testing, cutting lead time by about 50% and cost by about 30% versus the forged and welded version. Hard points are now printed directly into the tank wall, and a feed tube doubles as a weld locating point for the propulsion lines added afterward. None of it required a new alloy, just a decision to stop treating the tank as an assembly of parts.
As a critical piece of flight hardware, the printed tank is still subject to the same rigorous testing. A propellant tank is a pressure vessel, and pressure vessels going to space do not get a second chance. “For us right now, the technical hurdle remains the [non-destructive evaluation] NDE of these things,” Thompson said.
“A single piece tank becomes relatively difficult to inspect,” he said. “So the work that we’re focused on right now is what the next generation of NDE looks like, so this is something that is certified for spacecraft flight.” However, the part that solved the supply chain problem also raised a new certification challenge, a familiar tradeoff for any process moving from demonstration to flight qualification.
Honeycomb panels run into a version of the same tradeoff, though the complication shows up later. Switching to AM cut cost by about 90% on its own, and redesigning the internal lattice with topology optimization added more, trading mass for stiffness or the reverse, a 10% stiffness gain or a 15% mass cut depending on what the program needed. Much of the savings came from printing mounting hardware directly into the panel instead of bonding it on afterward.
So what was the complication? The aluminum alloy used so far, AlSi10Mg, does not survive heat treatment cleanly.
Asked to put a number on the distortion, Thompson explained, “What I can quantify is how much AlSi10Mg warped, like half of an inch when we went to quench it. We were shocked the part was still in one piece.” “It warped a lot,” he said. “You can see with the naked eye.” The parts looked fine off the build plate. The thin-walled geometry that made the panel light also made it unstable.
Northrop’s answer is CP1, an alloy Thompson said carries twice the electrical and thermal conductivity of AlSi10Mg and skips the quench step that caused the warping. He could not yet say by how much. “CP1 part is getting made in like a week, so I can’t quantify the difference,” he said. “CP1, 400° for four hours, stress relief,” he said. “It’s not going to have that same effect.”
Designing Backward From the Flight Product
Alongside tanks and panels, Thompson named other targets, among them solid rocket motor nozzles designed to eliminate mandrels entirely, integrated optical benches, and cable harnesses and ECS ducting cheap and fast enough to produce a part in hours rather than weeks. Thrusters, he added, are further along than people might assume. “I think the market understands the value of liquid rocket motors, some of the thrusters we use in satellites,” he said.
Mandrels are where the bottleneck looks different. Northrop uses AM to replace forgings used as molds for composite and metal structures, and the hurdle is not inspection but the cost of getting a new design approved.
“The challenges we have right now are expensive and long [non-recurring engineering] NRE,” Thompson said. He attributed it partly to DED’s immaturity compared to laser powder bed fusion (LPBF), which has had more time to automate the work before a part is printed.
Antennas are where this tradeoff has already been resolved, at least for one product. Feed horns built using LPBF in AlSi10Mg, developed with partner SWISSto12, are already flying.
“These are currently up in orbit doing their job,” he said. Printing the feed horn and feed chain as one part shrinks the assembly’s size, weight, and power draw. The detail Thompson kept returning to was surface finish. “Those surface improvement techniques are critical to achieve performance,” he said. “It’s really the big enabler for a lot of these products.”
Whether a part like that gets built at all depends on approval. Qualifying a new additive process still runs on the same slow, expensive cycle, generating test coupons against standards Northrop compiled into a document it calls the Space Additive Manufacturing Requirements Standard (SPAMRS), built by decomposing and tailoring existing frameworks, among them NASA-STD-6030, AWS D20.1, MMPDS, and AMS specifications, into something programs could apply consistently.
“This does not eliminate the data burden,” he said. “Generating data, generating coupons is still slow, it’s still expensive, waiting for someone to tell me how I can reduce it from 18 months and millions of dollars.” The same NDE problem that stalled the tank resurfaces here as a cost hurdle. “Generally speaking, quality makes up about half the part cost for additive parts today,” he said.
Some of that cost comes from duplicated effort, each company qualifying parts against its own standards rather than a shared one.
Thompson chairs an America Makes advisory group working on exactly that, alongside an internal push to release more material data Northrop no longer considers competitive. A related challenge is commodity materials like AlSi10Mg, where Thompson wants industry demand spread across more suppliers rather than locked to whichever one a program already qualified, so new machine capacity does not depend on a single vendor.
When asked for the one insight worth remembering, Thompson skipped past materials entirely. “Everyone always wants to talk about material systems,” he said. “But it forgets three other swim lanes that go into products, which is how do I qualify it? How do I design for it? And how do I inspect it?”
His method is to start from the product the company needs and work backward through qualification, design, and inspection until the additive work is “useful,” a low bar that a surprising number of additive projects still fail to clear.
That same caution showed up when he was asked about the US Department of Defense’s FY2026 budget. “I’m always going to be a proponent for more money,” he said. He added a qualifier that kept the talk honest. Additive manufacturing, he said, is “not the only answer” to the larger problems the department is trying to solve.
Grab your spot today for our AMA: Aerospace, Space, and Defense 2026 online conference on July 9th!
3D Printing Industry is inviting speakers for its 2026 Additive Manufacturing Applications (AMA) series, covering Energy, Healthcare, Automotive and Mobility, Aerospace, Space and Defense, and Software. Each online event focuses on real production deployments, qualification, and supply chain integration. Practitioners interested in contributing can complete the call for speakers form here.
To stay up to date with the latest 3D printing news, don’t forget to subscribe to the 3D Printing Industry newsletter or follow us on LinkedIn.
Explore the full Future of 3D Printing and Executive Survey series from 3D Printing Industry, featuring perspectives from CEOs, engineers, and industry leaders on the industrialization of additive manufacturing, 3D printing industry trends 2026, qualification, supply chains, and additive manufacturing industry analysis.
Featured image shows Additive Manufacturing Advantage: Aerospace, Space and Defense banner. Image by 3D Printing Industry.