Industry Watch · Additive Aerospace
NASA just bet a Mars mission on a 3D‑printed rocket company
Relativity Space will fly NASA's Aeolus atmospheric instruments to Mars in 2028. The "3D‑printed rocket" headline is real — but the most interesting part is where the printing stops, and what that teaches anyone who runs a printer.
Target launch year for the Aeolus mission
NASA-built science instruments aboard
Share of Terran 1 that was 3D printed, by mass
One Martian year of committed science ops
The Announcement
What NASA actually signed
On June 17, 2026, NASA announced a public‑private partnership with Relativity Space to carry a suite of atmospheric instruments to Mars. The split is clean: NASA supplies the science payload, and Relativity supplies the spacecraft, the rocket, and the cruise operations to get it there. NASA says the target is a 2028 launch.
The instrument package is called Aeolus — a fitting name, since the Greek Aeolus was keeper of the winds, and the mission's whole job is reading the Martian air. NASA describes Aeolus as the first effort to deliver an integrated, daily, global picture of Martian winds, temperatures, dust, and clouds all at once. The agency frames the payoff plainly: better atmospheric data means safer entry, descent, and landing for the robots — and eventually the crews — that follow.
The structure is unusual too. NASA calls it the agency's first six‑year reimbursable Space Act Agreement, a contracting mechanism that lets NASA work with a commercial partner outside a traditional procurement. NASA Administrator Jared Isaacman called partnerships like this "a force multiplier for science." NASA has committed to operate the instruments for at least one Martian year — roughly 687 Earth days — while Relativity maintains the spacecraft.
The Payload
Four instruments, one job: read the weather on Mars
The Aeolus payload is being designed, built, and integrated at NASA's Ames Research Center in California's Silicon Valley. According to NASA, it carries four complementary instruments:
Doppler Wind and Temperature Sounder (DWTS‑Ozone) — measures wind and temperature profiles from the surface up to about 37 miles (60 km), built in collaboration with GATS.
Thermal Limb Sounder (TLS) — builds vertical temperature profiles and watches dust and water‑ice clouds, developed with Xiomas Technologies.
Surface Radiometric Sensor Package (SuRSeP) — tracks how the Martian surface absorbs, stores, and releases heat, plus the properties of dust and clouds.
Wide‑Field Context Camera (WFCC) — takes daily global images to map atmospheric activity across the whole planet.
NASA positions Aeolus as the next step beyond two decades of orbiters like MAVEN, the Mars Reconnaissance Orbiter, and Mars Odyssey — not a lander, but an orbiter feeding a daily atmospheric model. The agency also says it will build a data‑processing pipeline to turn raw measurements into ready‑to‑use products for the wider research community.
Why The Headline Says "3D Printed"
Relativity earned the reputation the hard way
Relativity didn't get the "3D‑printed rocket" label from marketing. It got it from a genuinely radical premise we've been watching for years: build a rocket out of giant metal printers instead of a conventional supply chain of thousands of machined and welded parts.
The company's pathfinder, Terran 1, was about 85% 3D printed by mass — body and engines — using what Relativity calls wire‑arc additive manufacturing on its in‑house "Stargate" printers. In March 2023 it lifted off from Cape Canaveral, pushed through Max Q (the moment of peak aerodynamic pressure) intact, then suffered a second‑stage anomaly and didn't reach orbit. It never carried a paying payload. But it answered the one question skeptics cared about most: a printed pressure structure could survive the violence of launch without splitting along its layer lines. Standing about 110 feet tall, it remains one of the largest 3D‑printed objects ever to fly.
This is the same metal‑additive thread running through the rest of modern aerospace — the kind of work we cover in our piece on how metal 3D printing reshaped rocket propulsion. Printing wins where parts are complex, low‑volume, and benefit from consolidation — exactly the profile of a rocket engine.
The Part The Press Release Skips
The rocket flying Aeolus prints less than Terran 1 did
Here's the nuance worth keeping. Aeolus won't fly on Terran 1. It's slated to fly on Terran R — a much larger, partially reusable vehicle Relativity pivoted to after retiring the smaller rocket. And Terran R, by the company's own account, leans far less on 3D printing than its predecessor.
According to reporting by Ars Technica (summarized across the additive‑manufacturing press), Relativity CEO Tim Ellis acknowledged he could no longer claim Terran R would be roughly 90% additively manufactured. The same reporting described printing difficulties at large scale — including a "large buckling event" on a printed pressure dome — and indicated Relativity would lean on more conventional construction, such as aluminum sheet for the fuselage, with some structures potentially sourced from established aerospace suppliers. We're attributing those specifics to that reporting rather than stating them as settled fact; Relativity has released limited technical detail publicly.
None of that is a scandal. It's engineering. Additive manufacturing is being used surgically — on engines and complex components where it clearly wins — and set aside for large, simple, load‑bearing shells where rolled metal is cheaper and more predictable. "3D‑printed rocket company" stays true; "entirely 3D‑printed rocket" was always going to bend under physics.
Reality check
There's real execution risk here, and it's fair to name it. As multiple outlets have noted, Terran R hasn't flown yet — its first flight is targeted for late 2026 — and Relativity has not completed an orbital mission. NASA hasn't disclosed the value of the agreement. A 2028 Mars window is unforgiving: planetary alignment doesn't reschedule. The mission depends on hardware that still has to prove itself. That's the honest frame — promising, ambitious, and unproven all at once.
The lesson isn't "print everything." It's "print the parts where printing wins, and have the discipline to use something else everywhere else."
// The print-where-it-wins principle, from the bench
Side By Side
Terran 1 vs. Terran R: how the additive role changed
| Terran 1 (pathfinder) | Terran R (Aeolus launcher) | |
|---|---|---|
| Status | Flew once, March 2023; retired | First flight targeted late 2026 |
| Reusability | Expendable | Partially reusable |
| 3D-printed share | ~85% by mass | Reduced; engines/complex parts emphasized |
| Fuselage approach | Printed structure | Reportedly more conventional (e.g. aluminum sheet) |
| Where AM still leads | Body + engines | Engines, ducting, consolidated complex parts |
| Role in this mission | Proof of concept | Carries NASA's Aeolus payload to Mars orbit |
Terran R figures reflect public reporting and company statements; Relativity has disclosed limited technical detail, and specifics may shift as the vehicle matures.
From The Bench
Why a San Diego print shop reads this story closely
We don't print rockets. We run a Bambu Lab A1 and a Creality CR‑10S on the FDM side and an Elegoo Saturn 4 Ultra 16K on resin, every day. And the exact failure modes that reportedly bit Relativity's biggest parts are the ones we manage on a much smaller scale, constantly.
Same physics, different size
Layer‑line cracking under load. A printed wall is strongest in‑plane and weakest across its layers. Orient a stressed part wrong and it splits exactly where the layers stack. Relativity's whole Terran 1 flight was, in part, a test of whether printed structure holds under pressure — the same question we answer every time we choose print orientation for a bracket that has to take a load.
Large‑format warping and buckling. The bigger the part, the more a tiny per‑layer error compounds into a real dimensional or stability problem — which is the whole reason large‑format printing is its own discipline. A "buckling event" on a giant printed dome and a warped corner lifting off a print bed are the same gremlin at different scales.
Knowing when not to print. The most valuable judgment in this whole field is recognizing the part that shouldn't be printed at all. That's not anti‑3D‑printing — it's what separates a finished part from a failed one.
There's an upside thread too, and it's the part we love. The reason Relativity bet on printing in the first place was iteration speed: change the file, print again, fly again. That advantage is just as real at desktop scale. When we do scan‑to‑print reverse engineering, printing one test revision costs dollars, not the hundreds a machine shop would quote — so we iterate until a replacement clicks into place like the factory part. And the front of that pipeline is increasingly software: AI‑driven design tools like the ones we covered in our PhysicsX piece generate geometry so complex that additive manufacturing is often the only practical way to make it.
The Takeaway
What this means if you make things
The Aeolus deal is a milestone for commercial spaceflight and for additive manufacturing's place in it — a major agency trusting a printing‑native company with an interplanetary payload. It's also a clear‑eyed reminder that 3D printing is a tool with a sweet spot, not a magic wand. The companies and shops that win with it are the ones that know exactly where the edge of that sweet spot is.
That's the same judgment we bring to every job on the bench: the right material, the right orientation, the right process — and the honesty to say when printing isn't the answer. Whether you're prototyping a product, replacing a part that no one sells anymore, or just trying to figure out if your idea should be printed, milled, or molded, that judgment is the service.
Got a part that needs to be made right?
FDM from $7/hr, resin from $9/hr, plus 3D scanning and reverse engineering — from our Carmel Valley shop, serving all of San Diego County.
📞 858-342-6984 · ✉️ dreaming3dprinting@gmail.com
📷 @dreaming3dprinting · 📍 Carmel Valley, San Diego
Questions
Aeolus & the Relativity Mars mission, answered
Is Relativity Space's rocket really 3D printed?
Partly, and the share has changed over time. Relativity's first rocket, Terran 1, was about 85% 3D printed by mass and reached space in 2023. The larger Terran R that's slated to fly Aeolus uses 3D printing more selectively — heavily on engines and complex parts, less on big structural shells, which reporting indicates use more conventional construction. So "3D-printed rocket company" is accurate, but Terran R is not an entirely printed vehicle.
What is the Aeolus mission and what will it study?
Aeolus is a NASA atmospheric-science payload of four instruments designed to give the first daily, global view of Martian winds, temperatures, dust, and clouds from orbit. NASA says the goal is to improve atmospheric models so future landings — robotic and eventually crewed — are safer and more predictable.
When is the launch, and is the date solid?
NASA is targeting a 2028 launch. Treat that as a target, not a guarantee: the rocket that will carry the payload, Terran R, has not yet flown — its first flight is aimed at late 2026 — and Mars launch windows are dictated by planetary alignment, so slips have real consequences. The schedule is ambitious by design.
Why would a rocket company stop printing whole rockets?
Because 3D printing has a sweet spot. It excels at complex, consolidated, lower-volume parts — like engines — where it removes assembly steps and weight. For large, simple, load-bearing shells, conventional rolled or welded metal is often cheaper and more predictable, and printed structures at that scale can run into cracking and buckling problems. Using each method where it's strongest is good engineering, not a retreat.
What's "wire-arc additive manufacturing"?
It's a metal 3D printing method that melts metal wire with an electric arc and deposits it bead by bead to build large parts fast — the approach behind Relativity's Stargate printers. It trades the fine resolution of powder-bed metal printing for speed and scale, which is why it suits rocket-sized structures rather than tiny precision parts.
Does any of this apply to desktop 3D printing?
More than you'd think. The headaches Relativity reportedly hit at rocket scale — layer-line weakness under load and large-format warping — are the same ones every FDM and resin user manages, just smaller. Print orientation, when to reinforce a part, and when not to print at all are everyday decisions on our bench, not just in aerospace.
Can Dreaming3D print functional parts for me in San Diego?
Yes. We provide FDM printing from $7/hr, resin printing from $9/hr, plus 3D scanning and scan-to-print reverse engineering, from our Carmel Valley shop serving all of San Diego County. Send photos of your part or your file and we'll tell you honestly whether printing is the right process — and how to orient and reinforce it so it survives real loads. Call or text 858-342-6984 or email dreaming3dprinting@gmail.com.
Sources & further reading: NASA news release on the Aeolus public-private partnership (nasa.gov, June 2026); reporting from Space.com, ExecutiveGov, and others on the Relativity Space agreement; Relativity Space materials and prior reporting (including Ars Technica) on Terran 1, Terran R, and Stargate printing. Mission specifics, the agreement's value, and Terran R's final design reflect the best available public information as of June 2026 and may change.
This article is independent commentary by Dreaming3D and is not affiliated with or endorsed by NASA or Relativity Space.