155 m
Distance from the Apollo 12 lunar module to Surveyor 3. The Surveyor was still sandblasted: ~103 pits/cm², and that was the edge of the plume. [3]
InfrastructureSolar regolith paving
Population Mars · open hardware R&D
Solar Regolith Paving
Melt coarse rock with concentrated sunlight, stir in preheated fines, and anneal the mush into a 5 cm crystalline road base: asphalt, made with sunlight, for landing pads and roads on the Moon and Mars.
Stage: V1 hardware in procurement. No test data yet.
- 5 cm target depth per pass. ESA PAVER: ~2 cm. Target
- 3–5 MJ/kg energy per kg placed. PAVER lab parameters: ~10–14 MJ/kg. Projection
- ~70 % filler the pool can absorb if the filler is preheated to 1,000 °C. Cold filler: ~30 %. Calculated
- 300–450 MPa compressive strength of industrial cast basalt, the target material class. Reference
01 · The problem
Every landing sandblasts the site.
Rocket exhaust hitting loose regolith throws dust and gravel at hundreds of metres per second. There are no pads and no roads on the Moon or Mars, and every future landing, rover route and habitat site needs them.
300–2,000 m/s
Estimated speed of regolith ejected by the Apollo 12 descent plume. Lunar escape velocity is ~2,400 m/s. [3]
5–18 cm
Soil eroded under the Phoenix lander on Mars, exposing subsurface ice. Cratering under a human-class Mars lander could threaten the mission. [4]
The baseline
ESA PAVER, 2023
The best published result: a laser standing in for a Fresnel lens melted lunar simulant into interlocking glass tiles.[1][2] It proved the idea. Its limits define our design.
Where it stops
- Slow: a 45 mm spot moving at 5 mm/min.
- Glassy, smooth top surface; crystalline only lower in the layer.
- Fine powder feed: a regolith dust bed insulates, so melt depth stalls near 2 cm.
- Re-heating an already solid track cracks it; tile geometry had to avoid crossovers.
- Team
- BAM, Aalen Univ., Clausthal Univ. of Technology, LIQUIFER, DLR; ESA-funded
- Heat source
- 12 kW CO₂ laser standing in for a ~2.37 m² Fresnel lens
- Feed
- EAC-1A lunar simulant, dried 3 days at 180 °C to stop bubbling
- Process
- 45 mm spot at 188 W/cm² (~3 kW), 5 mm/min
- Product
- Interlocking hollow-triangle tiles ~20–25 cm across
- Depth
- ~1.8–2 cm per layer; 2.5 cm max after a 40-minute stationary dwell
- Strength
- 94 MPa compressive, ~4.8 % porosity
- Estimate
- 100 m² pad at 2 cm thick in ~115 days (ESA)
02 · The thesis
Melt the rock, not the dust.
Do not melt everything to glass, and do not excavate and sieve aggregate. Melt pools in the coarse rock, stir medium and fine aggregate into the pool, and screed the result: a partially molten crystal mush, the way asphalt is aggregate held in a binder. The process runs across a wider temperature band than full melting, and the product is thicker and rougher.
Energy
Melting basalt from ambient takes about 2 MJ/kg in theory (sensible heat plus ~0.4–0.5 MJ/kg latent). In practice, full-melt processes spend several times that, because a 1,400 °C surface radiates ~440 kW/m² into vacuum. We melt only 25–30 % of the placed mass and bring the rest to ~1,000 °C, where radiative loss is a third as large.
σT⁴ at 1,400 °C ≈ 440 kW/m²; at 1,000 °C ≈ 150 kW/m² (ε = 1)
Fracture
Glass has no crack-arrest mechanism. Unmelted crystalline aggregate deflects and pins cracks. Industrial cast basalt, melted at ~1,280–1,300 °C and cooled slowly in kilns, recrystallizes to pyroxene and reaches 300–450 MPa compressive strength with Mohs hardness 8.
PAVER glass tiles: 94 MPa. Cast basalt: 300–450 MPa.
Seeding
Filler of the same composition acts as a nucleant. A seeded basaltic mush should crystallize faster and at less undercooling than a clean melt, giving a cast-basalt-like ceramic instead of glass. V1 tests this directly.
Hypothesis under test in V1 (experiment 04).
Heat transfer
A regolith dust bed in vacuum conducts about 0.001–0.01 W/m·K. Solid basalt conducts ~1.5–2.5 W/m·K, 100–1,000× more. Flux on cobbles couples deeper than flux on dust. Coarse feed also carries less adsorbed volatile per kg.
Lunar regolith: 0.00074 W/m·K at the surface (Diviner); 0.01–0.03 at depth (Apollo).
Sources: energy [7]; cast basalt [5, 6]; PAVER [1]; conductivity [8, 9].
PAVER full melt vs this approach
| ESA PAVER (measured / published) | Solar paving (design) | |
|---|---|---|
| Feed preparation | Fine simulant powder, oven-dried | Coarse rock windrowed in place; fines preheated and degassed on board. No sieving. Target |
| Mass melted | ~100 % | 25–30 %; remainder preheated filler Target |
| Energy per kg placed | ~10–14 MJ/kg of beam energy (our calculation from the lab track parameters) | 3–5 MJ/kg delivered Projection |
| Depth per pass | ~1.8–2 cm | ≥5 cm Target |
| Surface | Glassy, smooth | Grooved by the screed for traction Target |
| Microstructure | Glass on top, dendritic crystals below | Crystalline throughout (seeding + annealing hood) Target |
| Cracking | Cracks when a solid track is re-heated | Cracks directed into joints scored every 0.5–1 m Target |
| Placed volume per m² of collector per sunlit day | 0.003–0.004 m³ (lab parameters) · 0.007 m³ (ESA pad estimate) | 0.0065–0.011 m³ Projection |
How we got this number: 100 m² at 5 cm in 46–77 days of sunlight
- Collector aperture
- 10 m²
- Lunar beam irradiance
- 1,361 W/m²
- Optical efficiency (reflectance × intercept × tracking)
- 0.75 (assumed)
- Power delivered to the ground
- ≈ 10.2 kW
- Pad
- 10 × 10 m × 5 cm = 5 m³
- Product density
- 2.7 t/m³ (cast basalt 2.9–3.0, minus ~7 % porosity, assumed)
- Mass placed
- ≈ 13.5 t
- Ideal heat content of the product
- 0.3 × 2.1 MJ/kg (melt to 1,450 °C) + 0.7 × 1.0 MJ/kg (filler to 1,000 °C) ≈ 1.3 MJ/kg
- Assumed thermal efficiency
- 26–43 % → 3–5 MJ/kg delivered. PAVER’s lab parameters imply ~15–20 %.
13.5 t × 3–5 MJ/kg ÷ 10.2 kW ≈ 46–77 days of sunlight per 100 m² pad at 5 cm.
At an equatorial mare site the Sun is up ~14.8 of every 29.5 days, so ≈ 3–5 calendar months. Polar ridges with longer illumination shorten this.
PAVER basis: ~3 kW at 10–14 MJ/kg and ~2.8 t/m³ gives 0.007–0.009 m³/day from a ~2.37 m² lens (0.003–0.004 m³ per m² per day). ESA’s 115-day estimate (2 m³, same lens assumed) gives 0.007. We project 0.0065–0.011.
What this means. Against ESA’s own pad estimate, our volume rate per m² of collector is similar. The gains we are testing are depth per pass (5 cm vs 2 cm), a crystalline grooved surface, controlled cracking, and no fines preparation. Against the published lab parameters the rate gain is 2–4×.
03 · Where the physics pushes back
Four problems, four design responses.
Each component of the train exists because of one of these.
- 01
Cold filler freezes the pool.
A pool at 1,450 °C has ~0.5 MJ/kg of usable heat before a mare-basalt mush locks up near 1,150 °C. Cold filler needs ~1.2 MJ/kg to reach that. Stirring in cold filler caps it at ~30 % by mass.
Response. Preheat the filler to ~1,000 °C (~0.2 MJ/kg still needed) and the pool can take ~70 %, an asphalt-like aggregate ratio. The preheat/degas car and a two-zone flux footprint (annular preheat + concentrated core, shaped by the iris) follow from this.
- 02
Volatiles foam the melt.
Lunar fines hold solar-wind gases. Martian fines released ~2 wt % water when Curiosity heated them to 835 °C; oxychlorines release O₂ and HCl at ~200–600 °C; iron sulfates release SO₂ from ~500–550 °C, magnesium sulfate from ~780 °C; calcium sulfate is stable past 1,000 °C.
Response. Degas at 600–1,000 °C before the material reaches the pool. On Mars this step is required. PAVER oven-dried its simulant for 3 days to stop bubbling; we do it in-line.
- 03
Radiation quenches the surface to glass.
A 1,000 °C surface still radiates ~150 kW/m² to space. Cooling through the 1,100 → 900 °C crystallization window happens in seconds on a bare ribbon, which is why full-melt tiles come out glassy on top.
Response. A polished-metal hood trails the melt and reflects most of that radiation back, plus low afterheat from spillover flux. Target: cut net radiative loss ~80 % so the ribbon cools slowly and crystallizes. The LWIR camera measures the cooling rate; hood position is the actuator. It is sheet metal.
- 04
Thermal contraction cracks a continuous ribbon.
From a ~900 °C set point to lunar night is ~1,000 K. With cast basalt’s 8–9 × 10⁻⁶/K expansion, that is ~0.8 % free contraction against well under 0.1 % tensile strain capacity.
Response. Score contraction joints every 0.5–1 m while the mush is still soft, as with concrete control joints. Same-composition filler and matrix keep internal expansion mismatch small.
Sources: volatiles [16, 17, 18, 1]; expansion [5].
04 · The paving train (V3)
Five cars, one continuous process.
A terrestrial asphalt paving train, reimplemented with concentrated sunlight in place of hot bitumen. The iris aperture shapes the flux into a preheat annulus and a melt core, and blinks closed for clean pyrometer readings.
- 01
Grade
Ambient
Steel blade and rake windrow fist-sized rock into the lane and firm the subgrade. No sieving, minimal excavation.
- 02
Preheat / degas
600–1,000 °C
Reflective hood and spillover flux heat the windrow and the filler fines. Water, chlorine and sulfur species vent before they reach the pool.
- 03
Melt
Pool ~1,400–1,450 °C
The dish focuses on the coarse rock; cobbles melt into a pool. A chute meters preheated fines in to form a crystal mush of up to ~70 % solids.
- 04
Screed
Mush 1,000–1,200 °C
A boron-nitride paddle spreads and lightly compacts the mush behind the hot core, presses traction grooves, and scores a contraction joint every 0.5–1 m.
- 05
Anneal
1,100 → 900 °C, slow
A long polished hood keeps the ribbon hot through the crystallization window so it sets as ceramic, not glass. Wheels run outside the paved lane.
05 · Materials
Feedstock and tooling.
Test feedstock
- Columbia River basalt, crushed V1 bulk feed
Free or cheap in Washington. Tholeiitic basalt: a reasonable, lower-iron analogue for mare basalt. Lets us burn through hundreds of kilograms learning the process.
- LMS-1 (Exolith) Lunar mare validation
Mineral-based mare simulant: pyroxene, glass-rich basalt, anorthosite, olivine, ilmenite. The main lunar validation feed.
- LHS-1 (Exolith) Lunar highland check
Anorthosite-rich highland simulant. Tests how far the process degrades on the harder-to-melt terrain.
- MGS-1 (Exolith) Mars validation
Rocknest-based Mars global simulant. Tests degassing and foaming with Mars-like chemistry.
On the Moon and Mars
- Mare basalt (Moon)
FeO- and TiO₂-rich; Apollo 12 basalt starts crystallizing at ~1,230 °C. Melts much more easily than highland rock. First choice.
- Highland anorthosite (Moon)
Plagioclase-rich; anorthite melts at ~1,550 °C. Workable but needs more flux per kg.
- Mare regolith grain size
Median grain ~0.07 mm; 95 % finer than ~1.4 mm. Coarse cobbles concentrate around fresh craters, so the demo site is mare next to an ejecta field.
- Mars rock cover
Modal rock cover ~6 % of surface area from orbital thermal data; no 1° × 1° region is rock-free. Landing sites measured 5–40 %. Coarse aggregate is easier to find on Mars than on the Moon.
The screed blade
The paddle works the 1,000–1,200 °C mush behind the hot core, the way an asphalt screed works behind the burner, never the 1,450 °C pool.
- Hexagonal boron nitride (h-BN)
Not wetted by most molten glasses and metals; standard release coating for glass forming and foundry work. Rated ~900 °C in air, ~1,400 °C in vacuum. Low mechanical strength and needs renewal, so the paddle has replaceable edges.
- Flight stack (Moon): BN on molybdenum / TZM
Refractory metal holds shape at screed temperature in vacuum. For Mars CO₂, the substrate choice is open: molybdenum oxidizes, and BN forms and loses a B₂O₃ layer in oxidizing gas.
- V1: BN-coated steel
Cheap coupons. In Earth air at 1,000–1,200 °C the BN coating is above its air rating, so it is re-sprayed each run and blade wear is logged.
- Not graphite
Carbon reduces FeO in the melt (gas bubbles, metallic iron) and burns in hot CO₂.
Sources: simulants [22, 23]; mare basalt [11]; grain size [10]; Mars rocks [12, 13]; boron nitride [20, 19, 21].
06 · Moon and Mars
Moon first, Mars second.
The Moon is the easier first deployment: full-strength sunlight, vacuum, and well-characterized mare basalt. Lunar landing pads are also where agency construction money is going now. Mars is the adaptation: larger collectors, sealed optics and mandatory degassing, but more coarse rock lying on the surface. The process physics is the same, so every lunar result carries over, and human-class Mars landers will need pads more than anything landing today.
| Moon | Mars | |
|---|---|---|
| Beam irradiance | 1,361 W/m², no atmosphere | ~590 W/m² above the atmosphere; ~200–400 W/m² direct beam at noon on clear sols (τ ≈ 0.4–0.9) |
| Collector area for the same melt power | 1× (10 m² in the projection) | ~3–7× at noon on a clear sol; more over a full day |
| Sunlight duty cycle | ~14.8 days on, ~14.8 off (equator); longer at polar ridges | Several usable hours per sol; weeks of downtime in dust storms (τ reached 10.8 in 2018) |
| Atmosphere and dust | Vacuum: good for melting, no oxidation; abrasive, electrostatically charged dust | ~6 mbar CO₂: oxidizes BN and metals; dust settles on and abrades optics |
| Coarse aggregate | Scarce in mare regolith; concentrated in crater ejecta | Common: modal rock cover ~6 %, 5–40 % at landing sites |
| Volatiles to remove | Solar-wind implanted gases in fines | ~2 wt % water, perchlorates/chlorates, sulfates |
| Verdict | Easier first deployment | Adaptation: bigger collector, sealed optics, mandatory degassing |
Mars beam irradiance is our estimate from Beer’s law, 590 × e−τ at the zenith [14]; storms [15]; rocks [12, 13]; volatiles [16, 17].
07 · Roadmap
Three stages, each with an exit test.
Budget sets the order: prove the material on Earth, fly it as a payload on someone else’s rover, then build the train.
Earth control head + solar concentrator
Build
- ~1 m² Fresnel concentrator on a 4-DOF head
- Computer-controlled iris aperture, 300 mm–1 m, on LinuxCNC + Python
- Blink-and-read: close the iris ~50 ms, read true thermal emission with no reflected sunlight, reopen — synchronized in LinuxCNC
- Two-colour ratio pyrometer (emissivity swings between dust, melt and slag)
- RGB video of the pool; LWIR camera on the wake to measure cooling rate
- Load cell on a manual rake: mush force vs temperature vs solids fraction
- Trays of crushed Columbia River basalt; LMS-1, LHS-1, MGS-1 for validation runs; argon tent as the middle step
Proves
The process window: melt depth in coarse vs fine beds, filler limit with preheat, foaming with and without degassing, and whether a hood turns glass into ceramic.
Exit criterion
Coarse beds melt ≥1.5× deeper than fine beds at equal energy; preheated filler reaches ≥60 % before lockup; hooded coupons are crystalline with cracks confined to scored joints.
Needs
V1 budget $15k–40k (estimate). Optics and ceramics reviewers. Coupon characterization (XRD/SEM). Test days in clear, dry sun east of the Cascades.
Paving payload on someone else’s rover
Build
- Melt head, preheat zone and annealing hood packaged as a payload
- Mounted on a partner’s mobility platform (small to large lunar rover class)
- Closed-loop control: iris, travel speed, filler rate and hood position from pyrometry and LWIR
- Vacuum or argon campaign at a partner facility
Proves
Continuous ribbon, not coupons: steady depth, joint spacing and cooling rate while moving; mass, power and dust behaviour of a flyable payload.
Exit criterion
A ≥10 m continuous ribbon at target depth in inert gas or vacuum, crystalline, with measured energy per kg within the projection.
Needs
Estimate $0.3M–1.5M: SBIR/STTR Phase I–II scale. A rover partner. Vacuum chamber or solar furnace time.
The full paving train
Build
- Five cars: grade, preheat/degas, melt, screed, anneal
- Blading and grading of in-place rock
- Deployable faceted dish with iris; sealed optics for Mars
Proves
Landing pads and haul roads built from local rock with no binder, at a rate a surface base can use.
Exit criterion
A 10 × 10 m pad at ≥5 cm in an analogue field test, then a flight opportunity.
Needs
Estimate $5M+, partner-led: agency programme or a commercial lander/rover company.
Cost figures are estimates and will be revised as quotes come in. V1 on Earth: a ~1 m² Fresnel lens delivers roughly 0.7 kW in clear, dry sun, enough for 2–3 cm melt pools. V1 studies the process, not throughput.
08 · V1 experiment matrix
What V1 measures, and what would stop the project.
Thresholds were set before testing. A kill result ends or redirects that line of work, and gets published in the log.
| # | Experiment | Variable | Measurement | Confirms | Kills |
|---|---|---|---|---|---|
| 01 | Melt depth vs bed | Flux (iris) × dwell × bed: fines <1 mm vs chips 5–30 mm | Sectioned melt depth; LWIR surface history | Coarse bed ≥1.5× deeper than fines at equal energy | ≤1.2× gain: the heat-transfer argument fails; the approach reverts to PAVER-like depth |
| 02 | Foaming vs preheat | Preheat: none / 600 / 800 / 1,000 °C; basalt, LMS-1, MGS-1 | Coupon porosity, bubble count, mass loss, video | Preheat ≥800 °C brings porosity near the unfoamed basalt baseline | MGS-1 still foams after 1,000 °C preheat: Mars needs a separate calcining step (Moon unaffected) |
| 03 | Filler loading limit | Filler fraction 0–80 % × filler at ambient vs ~1,000 °C | Lockup point: rake force spike + pyrometer | Cold ≈30 %, preheated ≥60 % (theory ~70 %) | Preheated filler locks up below ~40 %: the energy advantage roughly halves |
| 04 | Cooling rate → microstructure | Hood on / off; afterheat level | LWIR cooling rate through 1,100 → 900 °C; crack census; hardness; XRD on selected coupons | Hooded coupons crystalline, fewer cracks, harder than bare | No practical hood setting avoids glass: product is PAVER-like and the fracture argument fails |
| 05 | Mush rheology | Temperature × solids fraction | Rake load cell force | A workable window ≥100 °C wide at ≥50 % solids | Window <~30 °C: blading is impractical; V3 changes to pour-and-press |
| 06 | Coupon properties | Mush coupons vs full-melt glass controls from the same rig | 3-point bend, tumble abrasion, LN₂ ↔ kiln thermal cycling | Mush coupons beat the glass controls in bend strength and cycling | Mush coupons weaker than glass controls |
| 07 | Blink-and-read pyrometry | Iris closed ~50 ms vs open | Pyrometer vs thermocouple; surface decay during the blink | Reflected-sun error removed; decay correction small and repeatable | Engineering check, not a kill test: if the iris is too slow, add a small shutter at the sensor |
Every run is logged with the experiment-record template ↓ (bed type, preheat and filler temperature, filler fraction, hood state, LWIR cooling rate, rake force).
09 · Risks
Known risks.
Throughput
Even at the projected rate, a 10 m² collector places one 100 m² pad per 1.5–2.5 months of lunar sunlight. This builds pads, aprons and short haul roads, not highways.
Beachhead is dust-mitigation landing pads. Rate scales with collector area and the number of heads.
Air vs vacuum
In Earth air, FeO oxidizes, changing melt viscosity and crystallization. V1 results in air are process learning, not flight data.
Argon tent next, then vacuum at a partner facility (V2).
Mars dust on optics
Deposition and abrasion reduce concentration; global storms stop work for weeks.
Stow and cover optics; plan Mars duty cycle around storm season. Moon first.
Blade life
BN coatings are soft and oxidize in air and CO₂.
Replaceable edges; measure wear per metre in V1 and V2.
Thermal cycling
The lunar surface swings ~300 K between day and night. Joints handle the first contraction; long-term fatigue is unknown.
LN₂ ↔ kiln cycling on V1 coupons; longer campaigns at a partner.
Simulant fidelity
Columbia River basalt has less iron than mare basalt; simulants are not regolith.
Validate every key result on LMS-1 and MGS-1.
Team size
One person today, across optics, materials, controls and robotics.
The collaborator list below is the mitigation.
10 · Collaborate or fund
Who we need, and what for.
The project crosses optics, ceramics, controls, robotics and planetary science. Each line below is a specific piece of work.
Optics / solar thermal
Flux-map the V1 spot; design the two-zone footprint (preheat annulus + melt core) and the off-axis flight dish; dust-tolerant optics.
Materials / ceramics
Set the cooling schedule; section coupons; XRD and SEM on 10–20 samples; interpret crystallization vs glass.
Controls / pyrometry
Review the two-colour pyrometry and blink-and-read timing; LWIR cooling-rate mapping; LinuxCNC/Python control loop.
Robotics
Screed and rake mechanisms, joint scorer, train kinematics, V2 payload integration.
Planetary science / geotechnics
Site selection (mare near ejecta, Mars rock cover), volatile budgets, simulant fidelity review.
Test facilities
Vacuum chambers, argon gloveboxes, solar furnaces (e.g. DLR Cologne, PROMES-CNRS Odeillo, Sandia NSTTF): host one melt campaign.
Rover and lander companies
V2 ride-along: a payload envelope (mass, power, volume) to design against.
Funders
V1 budget now; small-business or university partners for SBIR/STTR; programme sponsors for V2.
Funding by stage
| V1 · Earth test head | $15k–40k Estimate |
|---|---|
| V2 · ride-along payload | $0.3M–1.5M Estimate |
| V3 · paving train | $5M+, partner-led Estimate |
V1 covers optics and 4-DOF head, iris, two-colour pyrometer, LWIR camera, simulants, argon tent, BN coatings and coupon characterization.
Channels we are pursuing
- NASA SBIR/STTR 2026 BAA Appendix B includes LIVEI.1.S26B (surface dust mitigation) and LAND.1.S26B (regolith stability during landing). Phase I up to $225k. [26]
- NASA MMPACT (Marshall) Moon-to-Mars Planetary Autonomous Construction Technology: landing pads, roadways, berms, blast shields from regolith. [24]
- LSIC (Johns Hopkins APL) Lunar Surface Innovation Consortium, Excavation & Construction focus area: monthly meetings where NASA construction stakeholders meet. [25]
- ESA Open Space Innovation Platform Open calls for ideas; PAVER itself was ESA-funded. [27]
Collaborate Fund V1
To collaborate, email feedback@haidaa.com. Say which experiment or stage you can help with.
11 · Later research threads
Bricks and heat storage.
The project began as solar-sintered MGS-1 bricks. Paving comes first because it is one continuous process with no molds or handling. Two threads stay on the list and reuse V1 hardware:
- Bricks and pavers. Sintered or melt-cast blocks for walls, arches, berms and interlocking pavers; strength, joints and thermal cycling.
- Thermal battery. Brick enclosures of loose regolith charged with concentrated sunlight by day to deliver heat through the night (Martian nights reach about −80 °C).
12 · Research log
Research log.
Priority change: paving first
The project now leads with solar regolith paving. Bricks and heat storage move to later research threads.
Decision. The first product is a paved surface, not a brick. Paving is one continuous process with no molds or handling, and it has a clear first use: landing pads, aprons and short haul roads that stop rocket-plume ejecta and dust.
Why now. ESA’s PAVER project (2023) showed that melting fine simulant into glass tiles works but is slow, shallow (~2 cm), glassy and prone to cracking. Our approach melts coarse rock, stirs in preheated fines to form a crystal mush, and anneals it under a reflective hood. The physics case is on the project page.
What changes in V1.
- Feed changes from sieved MGS-1 powder to crushed Columbia River basalt in fine and coarse beds, with LMS-1, LHS-1 and MGS-1 kept for validation runs.
- New instruments: an LWIR camera on the cooling wake, a load cell on a manual rake, and the iris “blink-and-read” pyrometry mode.
- New variables: bed type, preheat temperature, filler fraction and filler temperature, hood on/off.
- The experiment record template now includes these fields.
Status. V1 hardware is in procurement. No test data yet.
Project Kickoff: Sintering MGS-1 Bricks
Goals and experimental setup for the regolith brick sintering project.
We’re starting a hands-on research program to close the loop from loose Martian regolith simulant to a working thermal-storage structure, using concentrated sunlight as the only energy input.
Initial plan:
- Feedstock characterization — sieve MGS-1 into particle-size fractions and record baseline mineralogy and moisture content.
- Mirror rig — build/calibrate a parabolic mirror mount that can hold a stable focal spot on a brick mold, and measure the achievable focal temperature across the day.
- First sinter attempts — vary compaction pressure and dwell time, and document each firing on video.
Results, videos, and data will be posted here as log entries.
References
References
- Ginés-Palomares, J.-C. et al. (2023). Laser melting manufacturing of large elements of lunar regolith simulant for paving on the Moon. Scientific Reports 13, 15593.
- ESA (2023). How to make roads on the Moon (PAVER project summary; 115-day pad estimate).
- Immer, C., Metzger, P. et al. (2011). Apollo 12 Lunar Module exhaust plume impingement on Lunar Surveyor III. Icarus 211, 1089–1102.
- Watkins, R. N., Metzger, P. T., Mehta, M. et al. (2021). Understanding and mitigating plume effects during powered descents on the Moon and Mars. Planetary Science Decadal Survey white paper.
- EUTIT s.r.o. Cast basalt catalogue sheet E-01 (density 2,900–3,000 kg/m³; compressive 300–450 MPa; flexural ≥45 MPa; CTE 8–9 × 10⁻⁶/K).
- EUTIT. Cast basalt production (melted at 1,280 °C, cast, recrystallized in kilns).
- Bouhifd, M. A. et al. (2007). Thermochemistry and melting properties of basalt. Contrib. Mineral. Petrol. 153, 689–698.
- Hayne, P. O. et al. (2017). Global regolith thermophysical properties of the Moon from the Diviner Lunar Radiometer Experiment. JGR Planets 122.
- Mathew, N., Durga Prasad, K. et al. (2025). Thermal conductivity of high-latitude lunar regolith measured by ChaSTE on Chandrayaan-3 (incl. Apollo comparison). Scientific Reports.
- Carrier, W. D. (2003/2005). Geotechnical properties of lunar soil (median grain size ~0.072 mm).
- Experimental petrology of Apollo 12 basalts, part 1: sample 12009 (1971). Earth Planet. Sci. Lett. (olivine crystallizes from ~1,230 °C).
- Christensen, P. R. (1986). The spatial distribution of rocks on Mars. Icarus 68, 217–238 (modal rock abundance 6 %).
- Golombek, M. et al. (2003). Rock size-frequency distributions on Mars and implications for MER landing safety (5–40 % at landing sites).
- Appelbaum, J. & Flood, D. (1989). Solar radiation on Mars. NASA TM-102299 (mean 590 W/m²; beam irradiance vs optical depth).
- NASA JPL (2018). Opportunity hunkers down during dust storm (τ = 10.8).
- NASA (2013). Curiosity’s SAM instrument finds water and more in surface sample (~2 wt % water; heated to 835 °C). Leshin et al., Science 341.
- Review of SAM evolved-gas laboratory analog work: oxychlorine O₂ release ~200–600 °C, HCl ~200–542 °C. Minerals 11, 475 (2021).
- Lewis, J. M. T. et al. (2015). Sulfate minerals: a problem for the detection of organic compounds on Mars? Astrobiology 15 (SO₂ onset: Fe sulfates ~500–550 °C, MgSO₄ ~780 °C, CaSO₄ >1,000 °C).
- 3M Boron Nitride Suspension Cast-C datasheet (application temperature 900 °C air, 1,400 °C vacuum).
- AZoM. Boron nitride properties (h-BN not wetted by most molten metals, glasses and salts; coatings for glass-forming moulds).
- Jacobson, N. et al. (1999). High-temperature oxidation of boron nitride: I, monolithic boron nitride. NASA/J. Am. Ceram. Soc.
- UCF CLASS / Exolith Lab. LMS-1 lunar mare simulant (and LHS-1 highland simulant).
- UCF CLASS / Exolith Lab. MGS-1 Mars global simulant.
- NASA (2022). Moon-to-Mars Planetary Autonomous Construction Technology project: overview and status.
- LSIC Excavation & Construction focus area (Johns Hopkins APL).
- NASA 2026–2027 SBIR/STTR BAA, Appendix B (26B) SBIR subtopics.
- ESA Open Space Innovation Platform.
Solar regolith paving
Help build V1.
The next milestone is the first melt pools in fine and coarse basalt beds. Optics, ceramics and test-facility partners shorten that most.