3D Printed Solar Towers Can Become a Solution for Lunar Energy

Near the lunar poles the sun sits low on the horizon. A flat panel array collects very little energy at that angle, and the terrain shadows it for much of the rotation. Raising the collector to 50 m puts it above the surrounding ridges and into light that a ground-level array never sees. The published concept is a tall mast with a folding photovoltaic array near the top, roughly the shape of a highway traffic panel mounted on a column.

The tower is not only a power plant. It doubles as a communications mast, which matters because line-of-sight to a lander or rover is a recurring problem on a curved, cratered surface. One structure serving two functions reduces the number of separate deployments a mission has to land and assemble.

Foster + Partners developed the study with NASA and US technology partners under the Small Business Innovation Research program. A 1:1 build has not flown. What exists is a design study, scale models, and printed structural demonstrations shown at the Kennedy Center exhibition From Earth to Space and Return. Treat the numbers below as design intent, not flight heritage.

• 1
  Angle problemLow sun near the poles favors a raised collector over a flat array.
• 2
  Dual useSame mast carries power and communications hardware.
• 3
  StatusResearch study and exhibition models, not a flown asset.

Every kilogram launched costs far more than the same kilogram manufactured on site. That single fact drives the tower toward additive construction. The feedstock is lunar regolith, the dust and crushed rock already covering the surface. A printing head lays down regolith mixed with a binder, or sintered with concentrated solar energy or microwaves, in horizontal passes that build the mast upward.

The structural logic is a hollow tapered tube rather than a solid column. A shell carries bending load efficiently, and the hollow core leaves room for cable runs and a small elevator or climbing mechanism. Wall thickness is the main trade variable: thicker walls resist buckling, but each extra millimeter costs printing time and binder mass.

Printed regolith is weak in tension and brittle. It performs acceptably in compression, which suits a mast that mostly carries its own weight plus wind-free lunar gravity. Joints, the panel hinge, and the antenna mount are the weak points, and those are the parts most likely to be brought from Earth in metal. A hybrid structure of printed shell plus machined metal nodes is the realistic build path.

• 1
  FeedstockRegolith plus binder, or sintered in place.
• 2
  SectionHollow tapered shell, not a solid column.
• 3
  Weak pointsHinges and mounts stay metal, brought from Earth.

Lunar gravity is about one sixth of Earth's, so a 50 m mast does not need the mass a terrestrial tower of the same height would. The governing load is not gravity. It is the combination of a slow tip deflection over the array's swept area and the thermal cycling between sunlit and shadowed faces of the tube.

Thermal cycling matters more than most first-pass models assume. One side of the mast bakes while the other radiates to a cold sky, and that gradient bends the column a little every lunar day. Over thousands of cycles the printed shell can accumulate microcracks. A design with a machined metal spine or tensioned cable stays tolerates that cycling better than a bare printed tube.

For a crewed base the tower also has to survive launch vibration if it is brought assembled, or survive robotic assembly if it is not. Those two paths lead to very different joint designs. Robotic assembly favors simple, self-aligning interfaces with generous chamfers. Launch favors a folding mast with fewer, stronger hinges. The choice is usually made before the wall thickness is fixed.

• 1
  GravityRoughly one sixth of Earth; rarely the limiting case.
• 2
  ThermalDay-night gradient bends the tube and drives fatigue.
• 3
  AssemblyRobotic joints and launch joints pull in opposite directions.

Even a fully printed mast needs precision hardware. The panel hinges, the array deployment mechanism, the antenna gimbal, and the interface plate that ties the printed shell to the payload all fall into the machined-metal category. These are the components where a few hundredths of a millimeter changes how the array tracks the sun.

For ground prototypes we machine these in aluminum and stainless. A deployment hinge in 7075-T6 holds its geometry through repeated actuation. A gimbal yoke in 17-4PH stainless resists wear on the bearing seats. Test fixtures that hold a printed shell section during a bending test are usually 6061 plate, cut and pocketed to keep the fixture stiff without adding mass.

Tolerances on these parts are tight but not exotic. Hinge pin bores at ±0.005 mm give a repeatable array angle. Bearing seats held to Ra 0.8–1.6 μm seat a press-fit without galling. If a prototype array misaligns by a degree, the tracking loss shows up immediately in the power budget, so the hinge geometry is worth the extra inspection.

• 1
  Hinges7075-T6 for repeated actuation without wear.
• 2
  Gimbals17-4PH stainless on bearing and wear surfaces.
• 3
  Fixtures6061 plate, pocketed for stiffness at low mass.

A tower program moves through the same sequence as any hardware development. Design the node, check the print or cast route for the shell, then machine the metal interface parts so you can load them. The printed shell and the machined node have to fit on the first assembly, which means both sides need a shared datum and a tolerance stack that accounts for print shrinkage.

Rapid prototyping shortens the loop. A hinge body can be printed in a polymer to check range of motion, then cut in aluminum for load testing. Sheet metal fabrication covers the panel backing and mounting brackets. When the design stabilizes, small runs of the same parts come off the same machines, so the geometry does not change between the prototype and the test article.

No minimum order quantity helps here. A program that needs one hinge for a fit check and twelve for a vibration test can order both without a tooling investment. Uploads stay confidential, and an NDA is available when the design is not yet public.

• 1
  Shared datumPrint and machined sides must reference the same origin.
• 2
  ShrinkageBudget for print shrinkage in the tolerance stack.
• 3
  RunsOne part or twelve, same process, no tooling.

The published study shows a shape and a method. It does not show a power curve measured on the Moon. Solar cell efficiency at the lunar surface, dust accumulation on the array, and the actual regolith composition at the chosen site all move the output number by large margins. Any engineer sizing a base should treat the tower as one option in a trade study, not a settled answer.

Printing speed is the other open question. A 50 m mast printed in horizontal passes takes a long time, and that time is measured in Earth days of machine operation on a surface where maintenance is hard. Reducing height, or printing a shorter mast and using a cable-stayed array, changes the schedule a lot.

None of this makes the idea weak. It makes it early. The parts you can machine and test now, the hinges, nodes, and fixtures, are the same parts any future version of the tower will need.

• 1
  UnknownReal power output at the chosen site is not measured.
• 2
  SchedulePrint time for a 50 m mast is a program risk.
• 3
  CertainHinges, nodes, and fixtures are needed in any version.