Launching a kilogram to low Earth orbit costs in the low thousands of dollars even on the cheapest vehicles, and every kilogram of spare parts is a kilogram of payload you do not carry. That is why additive manufacturing became a serious space technology rather than a demonstration: if a crew can print a bracket, a tool, or a fluid fitting on demand, the mission stops carrying a warehouse.

Printing on a crewed station

Polymer extrusion printers have operated aboard the International Space Station for years. They print much as a desktop machine does, with the difference that the process had to be requalified for a place where you cannot vent fumes, cannot let a loose filament fragment float into someone's eye, and cannot send a technician up to unjam the nozzle.

Microgravity turns out to be kind to fused filament printing: the molten polymer is held by surface tension and the layer beneath it, not by gravity. The real problems were elsewhere — enclosing the build volume, filtering volatile organic compounds and ultrafine particles out of a closed-loop atmosphere, managing thermal behavior without convection, since hot air does not rise and a heated part cools very differently, and designing for maintenance by a crew member in gloves working from a checklist.

What gets printed is unglamorous and useful: tool handles, brackets, containers, adapters, and experiment parts. Comparisons between parts printed in orbit and identical parts printed on the ground have found broadly similar mechanical performance, which is what made the idea credible. Recycling has been demonstrated too — grinding used parts and packaging foam back into feedstock — which matters far more in orbit.

Manufacturing in vacuum, not just in a station

The more ambitious idea is building outside the pressurized module. In hard vacuum you cannot use most photopolymers, thermal management is radiative only, and outgassing contaminates optics — mission-ending for a telescope. But the payoff is enormous, because a structure built in orbit never has to survive launch.

That constraint drives an astonishing amount of spacecraft mass. Solar arrays, antennas, and booms are folded, hinged, and over-built mainly to survive a few minutes of vibration inside a fairing, then deployed once. A truss extruded in orbit can be sized for the microgravity loads it will actually see. Demonstrations have flown; economical operational systems have not.

Building with what is already there

Beyond Earth orbit the cheapest material is the material already on site. Lunar and Martian construction concepts center on binding regolith — the fine, abrasive, glass-sharp dust covering both bodies — into structural elements using sintering, microwave heating, or a small quantity of imported binder.

Testing with simulant material has produced usable bricks and printed structures, targeting applications where mass dominates: radiation shielding, landing pads that stop exhaust from sandblasting the base, and habitat shells. The hard problems are not the printing. They are power, since sintering is enormously energy-hungry; the abrasiveness of regolith on any moving mechanism; thermal cycling across a two-week lunar day and night; and the fact that a habitat wall must hold pressure, which turns construction into a pressure-vessel problem.

The part that is already routine: printing for space, on Earth

The most commercially significant additive manufacturing in the space industry happens in factories on the ground. Printed rocket engine components are standard practice across the launch industry: an injector or combustion chamber that once required dozens of machined and brazed pieces gets printed as one part with internal regenerative cooling channels that cannot be produced any other way. Fewer joints means fewer leak paths and fewer failure modes.

Satellite structures follow the same logic. Topology-optimized brackets, printed waveguides, and lattice heat exchangers all exploit three advantages: mass reduction where every gram carries a launch cost, part consolidation, and geometry machining cannot reach. At these volumes and mass penalties, a process that would be absurdly expensive for a consumer product is obviously correct — the same reasoning as metal 3D printing: when it makes sense and what it costs, taken to an extreme.

Qualification is the whole game

Here is the part worth studying regardless of your industry. Space hardware cannot be repaired, so the aerospace community had to invent a framework for trusting an additively manufactured structure.

  • The process is the specification. Machine, powder lot, layer thickness, laser parameters, build plate position, and orientation are frozen; changing any of them means requalifying, because they all change the resulting material.
  • Orientation is a design input, not a shop-floor choice. Additive parts are anisotropic, so flight parts carry the build orientation on the drawing — the principle in print orientation and layer adhesion, with far more rigor.
  • Witness coupons every build. Specimens printed alongside the part, from the same powder and run, are destructively tested to prove the build matched the qualified baseline.
  • Non-destructive inspection. CT scanning to find internal porosity and lack-of-fusion defects no surface inspection would catch.
  • Powder lot control. Reused powder degrades, so recycle counts, sieving, and oxygen content are tracked like a pharmaceutical ingredient.
  • Statistically derived allowables instead of a single datasheet number, feeding a conservative factor of safety.

What an earthbound product team should take from this

You are not going to CT-scan every bracket. But the underlying idea scales down, and most disappointments with printed production parts trace back to ignoring it: a printed part is defined by its process, not just its geometry. If you move a part from one service bureau to another, or the bureau changes machines or material suppliers, you have quietly changed the material properties of your product. Teams that treat a printed part like a machined part — send the file anywhere, get the same thing back — are the ones who find a strength problem in the field.

The practical version is modest: specify material, machine class, layer height, and orientation on the drawing; qualify a vendor rather than a quote; ask for lot traceability on structural parts; and test parts from the actual production process, not a nicely printed sample. That discipline is what makes the answer to are 3D printed parts strong enough a yes, and what separates printing end-use parts from printing prototypes. It feeds directly into reliability testing, and it is standard expectation in defense product development.

Projects House specifies printed parts the way structural parts should be specified — material, orientation, process, and acceptance criteria on the drawing — so production matches what you tested. If you are considering printed parts in a product that has to be trusted, describe the application through the contact form.