Sintered Regolith Roads and Machines in Lunar Dust

Published June 14, 2026

Shipping a cubic meter of inert mass from Earth means treating gravity as free. It is not. At current launch rates, that cubic meter costs millions in freight before doing any work. Local regolith brings its own headaches—power draw, heat, process control, and abrasive wear—but the dirt is already there. You do not pay Earth freight every time you need thermal cover or a blast barrier. The Moon does not need a city first; it needs to stop importing dumb mass for jobs local dirt can do.

• Pads
• Berms
• Shielding
• Haul routes
• Blast protection
• Thermal cover
• Crude structural mass

The business starts when local dirt takes over work that would otherwise chew through a launch manifest. The first real customer is not a tenant buying an Earth view. It is a logistics operator buying uptime: a landing provider wanting a reusable pad, an oxygen plant needing a berm, or a haul team tired of field repairs. On the Moon, construction begins as civil works, not real estate.

Nobody Pays First for a Moon House

The easiest way to get lost in lunar construction is to start with the habitat. Habitats matter, but they come late. Pretty interiors are dead last.

The real first market is classic industrial site prep: turning raw terrain into a workable yard. That means paving ground against rocket exhaust, piling berms around fuel lines, burying hardware under local dirt, and grading routes so every trip is not an expedition.

People keep asking what kinds of homes or factories we will build on the Moon. The better question: which site prep eliminates Earth imports fastest?

The answer is rarely pressurized volume. It is whatever bulky mass operators still launch because no local substitute exists yet. Follow the freight bill on bulk mass. When an operator launches inert shielding from Earth, gravity demands millions in launch costs before that material absorbs a single micrometeoroid or thermal swing. We know with certainty that shipping inert mass from Earth costs millions in freight before it does any work, while local dirt is already sitting on the surface. When local regolith takes over those heavy, simple jobs, the launch manifest carries revenue hardware rather than dumb ballast.

A landing operator does not need a cathedral. It needs a surface that cuts dust and ejecta.

A power system does not need a pavilion. It needs shielding, cable trenches, and less grit in its joints. A fuel depot does not need a campus. It needs graded pads, blast walls, and enough piled dirt to keep imported hardware alive. The Moon becomes buildable when local material takes over the heavy, simple jobs Earth-launched mass is too expensive to do.

There is a timing problem, though. The business does not start the moment one mission lands with a shovel. The economics change across operating scales. For a single demo where missions are short and landings rare, imported mats and disposable shielding win out because the launched mass is small. Once a site sees recurring landings and hardware stays in place, paying millions in freight to launch inert shielding across every tank, cable run, and storage yard breaks the budget. At that cadence, site reuse makes repair bills and freight cost more than moving local dirt.

The High Ground, Chapter 4: Built from Dust

In the European Space Agency's PAVER work, a 12-kilowatt carbon dioxide laser melts a spot of simulated lunar soil 4.5 centimeters across, then moves on and melts the next. The spots join into triangular tiles about 20 centimeters in size, and each pass leaves 1.8 centimeters of dense, fused material.

That matters because of what the alternative costs. The research behind this chapter puts a cubic meter of concrete or steel shipped from Earth to the lunar surface at $6 million to $19 million, at launch rates of $2,500 to $10,000 per kilogram. Sintering a cubic meter of lunar soil, 1,500 to 1,700 kilograms of it, with microwaves takes about 1,220 kilowatt-hours. Sintering means heating loose grains until they fuse into a solid without melting them all the way to liquid. The soil is regolith, the layer of broken rock and fine grains that covers the Moon.

The same research gives the road a second advantage. Ian Long's comparison puts iron at about 0.5 percent of the regolith mass that can be extracted, about 200 tonnes of regolith for each tonne of iron. A sintered road skips that refining step and uses the regolith directly. Regolith already spent by an oxygen plant can be sintered too.

Heat can reach the soil three ways: microwaves, concentrated sunlight, or a laser. The window that works is narrow. For EAC-1, a simulant made on Earth to match highland soil, the research finds a band only about 20 to 30 degrees Celsius wide, around 1,100 to 1,140 degrees, that brings the material to about 90 percent of its full density.

How strong the result is depends on which report you read. One gives compressive strengths of 40 to 80 megapascals for sintered regolith. The PAVER finding gives "several hundred" megapascals for its laser-melted tiles. The two figures come from different reports, and the research does not reconcile them.

Then the Moon heats and cools the work. A sintering report puts the swing at 300 kelvin, from 127 degrees Celsius in daytime to minus 173 degrees at night, once every 29.5 Earth days. A thermal report gives the same 300-kelvin-plus swing in its summary, then gives 120 degrees down to minus 130, a 253-kelvin swing, in its first finding. The research leaves those figures side by side. Either way, the sintering report treats that cycling as a design limit, and the mitigation it names is shapes that stay in compression, such as arches and domes. Those are the shapes whose material gets squeezed, not pulled apart, by the loads on them.

Apollo hardware ran into the dust first. Lunar dust grains are sharp, carry tiny particles of iron called nanophase iron, and cling to surfaces. On Apollo hardware, dust covering 11 percent of a surface doubled how much sunlight it absorbed as heat. The Apollo 12 magnetometer ran hot. So did the batteries on the Apollo 15 rover. The astronaut John Young called dust the number one concern in returning to the Moon.

Radiators, the panels that dump a machine's waste heat into space, suffer the same way. In a thermal report, dust covering 12 percent of a radiator raises the heat it absorbs by up to 50 percent, and dust covering 54 percent cuts how well it gives off heat by up to 16 percent. That report's conceptual south pole habitat has to reject 15 kilowatts by day through about 48 square meters of radiator, and needs 2,900 to 5,400 watts of heating to get through the night.

The book's outline calls dust mitigation a service industry. The buyers are the operators whose radiators, seals, and batteries stop working. Electrodynamic dust shields, which use electric fields to push dust off a surface, stand at technology readiness level 5 to 6, on the scale engineers use to rate how close a technology is to working hardware. Two questions stay open in the dust research. Dust in the permanently shadowed craters near the poles is described as distinct and untested, and the health risk to people is described as uncertain.

Moving parts face a second problem that has nothing to do with grit. In a vacuum, clean metal has no oxide film on its surface, and two clean metal surfaces pressed together can stick. This is called cold welding. In a test for the European Space Agency, 17-7 PH stainless steel coated with molybdenum disulfide, a dry lubricant, was rubbed against a titanium alloy, Ti6Al4V, in a vacuum below 10^-5 pascals. After about 40 fretting cycles, small back-and-forth rubbing motions, the coating failed and the force holding the surfaces together rose from 100 millinewtons to 6,000 millinewtons, a 60-fold increase. Bare pairs of pure metal stuck at more than 12,000 millinewtons. The vibration of launch alone can strip the coating before the part ever reaches the Moon.

That is the limit the outline names for this chapter: wear, heat rejection, and whether moving parts survive. A bearing that sticks stops whatever machine it sits in.

One fix is structural. The road research says a paved surface removes the loose regolith that the dust comes from, so a sintered road also controls dust.

On unpaved ground, wheels pay the cost. The lunar terrain vehicle is targeted at a 10-year life and 10,000 kilometers of driving. Lunar soil gives rolling resistance of about 0.3, and driving across it costs about 0.15 kilowatt-hours per tonne-kilometer.

Rail is the next step up. Northrop Grumman's 2024 lunar rail study, done under DARPA's LunA-10 program, describes trains carrying 100,000 to 125,000 kilograms per operation. The same research says rail is justified only at high throughput, 10 to 15 years into sustained operations.

FLOAT is a segmented track with no moving parts. Its carts raise no dust, it can climb 30 percent grades, and moving cargo on it costs about 0.02 to 0.05 kilowatt-hours per tonne-kilometer. Its maturation is stated as the mid-2030s, and joining the track segments together is still unsolved.

Cables go where wheels do not, over rough terrain and across shadowed ground, point to point, with no dust at all. Lunar gravity cuts the mass of cable and towers by about six times compared with Earth. For small loads there is the Micro Nova Hopper, which carries up to 10 kilograms about 25 kilometers from its lander.

A simplified cost model in the road research puts wheeled haulage at $0.023 per tonne-kilometer and rail at about $0.06. The report labels it simplified, and it is not a bid. The choice between wheels, rail, track, and cable comes down to how much mass moves along a route.

A different way to pull metal from rock is to let microbes do it, which is called biomining. In the BioRock experiment on the International Space Station, microbes working on basalt recovered up to 283 percent more vanadium in reduced gravity than at normal Earth gravity. A shared report on lunar biogeochemistry stops that result short. The test used basalt, at gram scale, and no published experiment has run microgravity biomining on lunar regolith or on a high-fidelity simulant. The outline treats it as a fork in the road, not the main route.

Under the surface there may be shelter. Data from the GRAIL and LRO spacecraft point to lava tubes, and one report states that tubes up to 4 to 5 kilometers wide would be stable under lunar gravity, and that a tube could be pressurized to 1 atmosphere with a roof as thin as 10 meters. One model by Martin uses a tube 120 meters across. The way in would be skylights, pits that open from the surface. Kaguya found the Marius Hills pit, given in the report as about 50 meters, in 2009, and others are known in Mare Tranquillitatis and Mare Ingenii. A thermal report says shaded spots in pits about 100 meters deep sit near 17 degrees Celsius.

The same subsurface report estimates that an oxygen plant on the surface would cost $2 billion to $3 billion and one in a lava tube $700 million to $900 million, 60 to 70 percent less. That is an estimate. The missions that would survey a tube, DAEDALUS and Spelunker, are concepts, not completed surveys.

Siting on the surface has its own tradeoffs. On the rim of Shackleton crater near the south pole, the thermal report gives 88.0 percent average sunlight over 20 years, so darkness lasts about 52 hours instead of 14 Earth days, and the energy storage needed drops from 1,680 kilowatt-hours at the equator to 260 kilowatt-hours.

The first lunar works in this research are pads, roads, and seals, built by machines that have to keep running in dust and vacuum. A cubic meter sintered from local soil costs about 1,220 kilowatt-hours, where a cubic meter of concrete or steel from Earth costs $6 million to $19 million. The sintering rig, the vehicle hauling the soil, and the bearings in both still face lubricant coatings that failed in testing after about 40 rubbing cycles, and dust shields that stand at technology readiness level 5 to 6. That is why the chapter's thesis puts abrasion control and transport geometry ahead of high-purity metallurgy as possible early markets. The same machines also have to get through the lunar night, and power is where The High Ground goes next. Open Chapter 5, Energy from the Sky.