In-space manufacturing and assembly

Every spacecraft ever flown was built on Earth and folded into a fairing. That constraint shapes designs so pervasively that it is easy to mistake for physics.

The fairing constraint

A spacecraft must survive launch loads and fit a cylindrical volume of at most about 5 m diameter for most vehicles, rising to 7 or 9 m for the newest. From Launch vehicles, that is the binding constraint for large lightweight structures far more often than mass is.

The consequences are visible in the hardware.

JWST folded. Its 6.5 m primary is 18 segments because a monolithic mirror would not fit, and its sunshield deploys from a package. That required 178 release mechanisms and produced 344 single-point failures, every one of which had to work. See Space telescopes.

Antennas and arrays deploy. Large reflectors and solar arrays are folded, which means hinges, latches and a deployment sequence that must succeed once, unattended, after launch vibration.

Structures are overbuilt. A structure sized to survive 8 g of launch acceleration and acoustic loading is enormously stronger than orbital operations require. Most of that strength is dead weight for the rest of the mission.

Assembling in orbit removes all three. A structure built in space needs no launch-load margin, no folding mechanisms and no fairing envelope, and the parts can be launched separately in whatever configuration packs best.

Robotic assembly and servicing

Hubble is the proof of concept for servicing value. Five crewed missions corrected the primary mirror’s figure error, replaced instruments four times, and repeatedly renewed the gyroscopes and batteries. The telescope’s scientific output is largely a consequence of being serviceable, and its most cited results came from instruments installed years after launch.

Mission Extension Vehicle demonstrated the commercial case. Rather than refuelling a satellite, it docks to it and provides attitude control and station keeping using its own propulsion, effectively becoming a replacement propulsion module. This is a clever simplification: it avoids the difficulty of transferring propellant into a spacecraft never designed to receive it. Two MEVs have extended Intelsat satellites’ operational lives.

The unsolved case is non-cooperative capture: a target with no docking interface, no markers and possibly tumbling. This is the enabling technology for active debris removal, discussed in Space debris and traffic management, and for servicing the large existing fleet of satellites that were never designed for it.

Additive manufacturing in microgravity

Printing in orbit has been demonstrated and the results are instructive about what actually matters.

Polymer printing works

Demonstrated on the ISS since 2014. Parts printed in orbit are comparable to ground-printed equivalents, which was the question being asked.

The value is logistics, not novelty

A printer replaces a spares inventory. Rather than launching every part that might fail, launch feedstock and print what breaks.

Recycling closes the loop further

A device that grinds failed parts back into feedstock reduces resupply mass again.

Metal printing is harder

Powder handling in microgravity, heat management without convection, and the power required all complicate it. Demonstrated at small scale.

Large structures are the real prize

Extruding beams and trusses in vacuum, unconstrained by fairing dimensions, is what would change spacecraft design rather than merely improve logistics.

The distinction between printing spares and printing structures is the important one. The first is a supply chain improvement with clear near-term value. The second is an architectural change that would allow apertures and arrays of a scale no launch vehicle can accommodate, and it remains a demonstration technology.

Manufacturing that requires orbit

A separate proposition is making things in space that cannot be made on Earth, so the product justifies the launch cost rather than merely tolerating it.

The honest position is that despite decades of interest, no product has yet established a business on this basis. Each candidate has a real physical rationale and none has cleared the combined hurdle of launch cost, return cost, production rate and a market willing to pay a premium.

The variable that changes the analysis is launch price. Every one of these propositions becomes more plausible as cost per kilogram falls, which ties this field directly to reusable launch. ZBLAN fibre is the frequently cited case because its potential value per kilogram is very high, but even it has not yet closed.

What would have to be true

For in-space manufacturing and assembly to become routine rather than demonstrative, several conditions must hold together.

Launch cost must be low enough that sending feedstock and machinery is cheaper than sending finished articles. Robotic manipulation must be reliable enough to work unsupervised, which means solving non-cooperative capture and precision assembly. And there must be a demand for structures larger than a fairing, which is currently limited to a handful of astronomy and power concepts.

The third condition is the one least often examined. Assembly capability without a customer for large structures is a solution waiting for a problem, and the strongest candidate customers, very large telescopes and space-based solar power, are themselves contingent propositions.

References

• Belvin, W. K. et al. “In-Space Structural Assembly: Applications and Technology”, AIAA 2016-2163.
• Lymer, J. et al. “Commercial Application of In-Space Assembly”, AIAA SPACE 2016.
• Prater, T. et al. “3D Printing in Zero G Technology Demonstration Mission”, International Journal of Advanced Manufacturing Technology, 101:391, 2019.
• Tucker, D. S. et al. “Elimination of crystallization in ZBLAN fibre processed in microgravity”, NASA technical reports on ZBLAN flight experiments.
• Reed, J. and Barnhard, G. “Satellite Servicing: The Mission Extension Vehicle experience”, AIAA ASCEND, 2021.