The first advanced lunar transportation concept is the idea of intentional hard lithobraking landings of bulk raw materials. The idea was recently suggested by frequent Selenian Boondocks commenter Paul Dietz in the comments to this previous post. Basically you would launch a rocket with one or more large chunks of raw material on a trajectory that grazes the surface of the Moon, using the lunar regolith to decelerate the raw material to the point where it can be subsequently collected. Even if the payload is coming in on a hyperbolic trajectory from Earth without stopping in low lunar orbit, it’s theoretically possible for a very shallow impact to not vaporize the payload, leaving it largely intact.
Artificial Shallow-Impact Meteorites
In Dennis Wingo’s excellent book MoonRush, he makes the argument that due to the Moon’s shallow gravity well, it’s possible for a subset of lunar nickel/iron meteorite impactors to arrive at low enough relative velocity and a shallow enough angle for the majority of the material to survive intact in a large chunk.
This concept though would be basically trying to create that phenomenon artificially, with an earth-launched chunk of raw material, such as a big ball of some alloy that may be hard to produce on the Moon (copper conductor alloys? high strength aluminum alloys?). Impact conditions could be carefully controlled using the upper stage as a shepherding vehicle, similar to what was done with the LCROSS mission1, but potentially with the upper stage performing a last minute maneuver to avoid intercepting the lunar surface, allowing it to return to earth for recovery/reuse.
Why would you want to do something like this?
Primarily because it would dramatically reduce the cost of delivering some bulk raw material to a lunar settelement site. The delta-V from LEO to a lunar soft-landing is approximately 6km/s. With high performance LOX/LH2 propulsion, that means that only 1/4 of your LEO mass will be dry mass on the lunar surface. For instance, an ACES stage with 70mT of prop loaded in LEO could deliver approximately 24mT to the lunar surface, with the stage and Xeus landing hardware taking up at least 6mT of that (leaving a net ~18mT of payload). On the other hand, the same stage could hard land over 65mT of bulk material while recovering the stage back to LEO. In the case of a big chunk of raw material, unless it is something really exotic, most of the cost is going to be in the launch of the material and the propellant to get it to the moon. If you can increase the payload to the moon for the same propellant by a factor of 3.6x, that’s likely going to decrease the cost on the lunar surface by a somewhat similar margin.
What would you use this for?
As mentioned above, the most likely use of this would be to hard land bulk raw materials. Specifically you’d want something somewhat tough so it could survive the impact intact, in spite of impact forces probably in the 1000s of Gs, most likely a large chunk or ball of metal. Here are some specific applications I can think of:
- Copper conductor materials — copper is relatively rare on the Moon, but is used a lot as a conductor for power transmission, and could be a key element in manufacturing electromagnets or conductors for other propellantless launch/landing systems. Landing equipment that could metal and process high-purity copper or copper alloys into useable conductors may be a lot easier than landing big spools of conductors, if this approach allows you to cut the cost of the copper by almost a factor of 3.6x.
- High-Strength Aluminum Alloys — while aluminum itself is pretty common on the Moon, most high-strength alloys, particularly ones you’d want to use for structural components or pressure vessels, are ones that require alloying elements that may be hard to source locally, and which frankly will likely take a long time to build up the smelting capabilities on the Moon even if the alloying elements are available. These alloys could be alloys that can be cast using locally produced sintered regolith molds. Or they could be alloys designed to be converted into 3D printing powder. Or it could be wrought alloys that are designed to be melted and converted into sheet or bar forms. Once again, processing equipment to convert the bulk raw material into usable forms is likely going to be massive enough that you’ll be somewhat limited, but with proper thought the 3.6x cheaper aluminum alloy raw material may still be useful in some applications.
- Toughness-Enhancing Alloys for Ni/Fe Asteroidal Material — in the ongoing discussion on the previously mentioned ESIL-8 post, it was mentioned that many examples of meteoric steel, while beautiful and stainless steel-like are actually rather brittle. It might be possible to send along the right mixture of alloying materials to mix with magnetically recovered Ni/Fe asteroid fragments on the Moon to turn it into a much better and more useful grade of stainless steel, for 3d printing, and/or production of pressure vessels for habitation or material storage.
Implementation Details
Some potential implementation issues that may need to be addressed for this concept include:
- What materials can actually survive an impact at the velocities associated with a hyperbolic trajectory from Earth?
- What is the optimal impact trajectory for controlled landing in a recoverable manner? Does too shallow result in large dispersions? Does too steep result in vaporization or the material being buried below the surface? How much of this can be analyzed up-front versus needing to be experimented with?
- How big of an impact crater are you going to form? How far will regolith fly?
- What is the optimal shape for the material? A solid sphere? A hollow sphere?
- From an earth TLI, where on the Moon can you realistically target an impact point for that has the right impact flight path angle?
- Where would you want material to be targeted relative to a settlement, to minimize risk to the settlement while simultaneously maximizing the odds of being able to recover the material easily?
- Can you design a trajectory that allows the shepherding upper stage to avoid lunar collision and return to Earth for recovery and refueling?
- Do you need to “prepare the ground” any by having a robot remove large boulders from the target area, and/or “rake” the regolith?
There are probably other questions, but the potential of being able to cut the cost of raw materials delivered to a lunar site by 3.6x makes this intriguing, even if a little on the messy side.
Next Up: Propellantless Soft-Landing Options
Jonathan Goff
Latest posts by Jonathan Goff (see all)
- NASA’s Selection of the Blue Moon Lander for Artemis V - May 25, 2023
- Fill ‘er Up: New AIAA Aerospace America Article on Propellant Depots - September 2, 2022
- Independent Perspectives on Cislunar Depotization - August 26, 2022
This is why mars is better for colonization. Why would you do this on the moon? Because it lacks certain elements a colony needs. This method wouldn’t work (unless encased?) for other required elements.
Add a hole and you’ve got the solar systems biggest game of golf!
Ken,
I think my point was more that even if the elements are there in the lunar regolith, having a way of getting large chunks down that don’t require tons of smelting, refining, and alloy mixing infrastructure might be a good idea. This would be true also on Mars, but it’s deeper gravity well makes this a lot less likely to work there.
~Jon
Would there actually be any benefit from a shallower angle if you want the object to stay in place at the impact sight? I can see the benefit in survivability from a grazing angle that makes the object skip off the Moon repeatedly but that would make recovery more complicated.
And I’m not sure vaporization beneath the surface is all that bad of an outcome if you’re just interested in the elemental copper/alloying agent and aren’t worried about ruining the temper of your aluminum.
Andrew,
There are definitely tradeoffs in impact angle vs survivability vs recoverability. The goal is to minimize the amount of equipment needed to recover the material and then turn it into useful end-products like wire or structural materials, etc. So you want to minimize contamination of as much of the material as possible, while also minimizing depth it gets to and minimizing travel distance to recover the material.
George Herbert suggested the idea of picking the trajectory to deliberately clip the lip of a mountain or crater. That can allow a lot of velocity to be bled off without a lot of burying.
Longer term, by the time you can afford to put lots of earth moving, excavating, and metal processing/refining equipment on the surface, there are probably better ways of getting bulk materials down.
~Jon
Pingback: Lunar Hard LandingsÂ
In my previous referenced comments I was imagining high angle impacts. Some materials will not (fully) vaporize even then, since they don’t have enough kinetic energy , even if we assume that kinetic energy goes 100% into the material and not the regolith. Lithium hydride, for example, has a heat of fusion (never mind vaporization, and never mind the energy needed to heat the material to the melting point) that is about equal to its kinetic energy at lunar escape speed. Under the same assumptions carbon would be heated to around its melting point.
The larger concern I’d have is digging up fragments (or recondensed vapors) if they get buried too deeply. This could be addressed by using very small impactors, which would also transfer heat more readily to the surrounding regolith. In the limit, this would mean dropping a stream of dust (or vapor!) that would form a crust on the surface that could just be scraped off.
Paul,
I’m not convinced that diffusing and mixing the delivered materials into the regolith is the best strategy, as that’s going to require a lot of work to collect and repurify the materials after the fact. While that might be an ok strategy for something the Moon really lacks, I don’t think it makes sense long-term as there are other propellantless landing methods that can get things down more gently, but just take a bit more infrastructure. As I mentioned to Andrew, I think this concept works best if done in a way that minimizes the amount of pre-landed hardware needed to take advantage of it. And to me, the main benefit is landing stuff that can be incorporated into lowering the cost of landing and launching future hardware.
~Jon
Selenian Boondocks is one of the better internet resources for momentum exchange tethers. Have thought of tethers as a way to mitigate high velocity impacts on the lunar surface?
Prograde lunar orbits are the most vulnerable to tidal damage from the earth while retrograde orbits are more durable. A polar orbit would be somewhere in between. A tether in polar orbit could drop payloads at the poles, the equator or any lunar latitude.
Deeper lunar orbits are are more resistant to damage from tidal influence. I believe orbits with apolunes beneath 40,000 km altitude will be long lived.
But go too low and lunar mascons will destabilize the orbit. However most of the orbits that have been wrecked by mascons are very low, on the order of 300 km altitude.
I believe a tether whose center of mass is above 10,000 km at perilune and below 40,000 km apolune will be long lived. The extremities of the tether can extend beyond those boundaries since the tether mass at the extremities is a small fraction of the over all mass and thus tidal or mascon perturbations on the tether extremities wouldn’t exert much force.
At the center of the tether can be an anchor mass. This would consist of solar panels, propellent and ion rocket engines. Ion engines can gradually build up a tether’s orbital momentum but the act of dropping or throwing a payload is quite sudden, even faster than an impulsive chemical burn. Thus tethers are a way to enjoy ion’s great ISP but still have a good Oberth benefit.
A tether’s momentum might also be preserved if traffic is two way. Catching from higher orbits and dropping to lower orbits boosts momentum while catching from lower orbits and throwing to higher orbits saps momentum. Balancing the boosts with hits can keep the tether in an orbit at a useful altitude without using a lot of reaction mass.
I’ve even seen some lunar rotovator schemes whose extremity touches the lunar surface while moving 0 km/s wrt moon. I believe this requires an unrealistic level of precision. But it might be doable to drop payloads to the lunar surface with an impact velocity less than 1 km/s.
That is a very interesting idea. I like it.
Hop David, momentum transfer tethers strike me as a very useful infrastructure at several locations. A lot of work to install, more so if you don’t have a convenient counterweight already in place. But they should pay off very well if a lot of payload is being transferred through a location.
Nice idea. A sphere would work best in an atmosphere, but for the Moon, my guess would be a penetrator. Long, narrow, and with a sharp point. It would go cleanly straight into the ground, and I think would be easy to recover. And would be contained, even if distorted / compressed all the material would end up in the hole it creates.. Small narrow penetrators, lots of them.
That would also deal with the boulders problem. If some of them hit a boulder no problem, if there are lots of them and you target an area with few boulders.
I’m glad there weren’t any people around to bet with. I would have bet that the hard landing was going to be into a cloud of electrostatic levitated dust considering Jons’ work in that field. 300 meters high for steep impact at 1,000 gee with margin. 3 km long for grazing stop at 100 gee.
Not completely crazy. There are some metals where this might be useful. However, there might be ways to slow this thing down without any retropropulsion. 🙂
Dennis,
Thanks for the comment! What do you mean by your last sentence though? The point of this idea was to enable recovering a payload landed without any post TLI retropropulsion (though maybe a tiny bit of steering to make sure you hit the right point on the moon).
~Jon
Jon
I know what you meant. In order to do something like this you are going to need a lot more than a little post TLI steering in order to make it cost effective. Hitting an exact point on the Moon is a non trivial activity. Off by a few km and you are in an inaccessible location, or goodby base….
What about spraying regolith dust up into the path of the incoming payload. The spray could be in increasing densities as the payload slowed down. The payload could be encased in a particularly hard sacrificial shell.
A variant on your concept would be to have a refuelable stage in LLO which docks with the payload and then decelerates the payload to a certain survivable velocity before detaching from the payload and then, with the reduced mass, heading back up to LLO to repeat the procedure. This wouldn’t be as cost-effective but it could ease the most challenging parts of the purist approach.
Cool idea.
Also, why not hollow out the hard material to enclose volatile material (e.g. rocket fuel) that would otherwise vaporize?
PeterH wrote “Hop David, momentum transfer tethers strike me as a very useful infrastructure at several locations. A lot of work to install, more so if you don’t have a convenient counterweight already in place. But they should pay off very well if a lot of payload is being transferred through a location.”
Just so.
There are possible sources of counterweights. There are many Near Earth Asteroids whose orbit energy with regard to earth is just above escape. High lunar orbit is just below escape. (See http://hopsblog-hop.blogspot.com/2015/05/eml2.html). The Keck proposal for asteroid retrieval points out some rocks can be parked in lunar orbit for as little as .17 km/s.
Another source of anchor mass might be the dead sats in the graveyard orbit just above geosynch. A vertical tether with a foot just above geosynch has some obvious uses. Most of these dead former GSO sats are traveling roughly the same velocity and at 0º inclination. It would be possible for an ion driven device to gather these into a single mass. Many of these dead sats have working solar arrays. If consolidated these solar panels could sum to a respectable power source.
Just to add a bit more on the reasoning for using penetrators:
I think a penetrator would decelerate more slowly than a sphere or any other shape. Would generate the same amount of heat, but over a longer time period giving an opportunity to dissipate some of it as it is generated, and also a chance to leave it behind if it penetrates several times its length into the regolith.. I think the best penetrators, the ones that travel furthest into the regolith without stopping, might reach lower temperatures than any other shape. So less likely to melt, and especially, not so likely to evaporate.
You could also use a crushable nose like a crumple zone to help decelerate, as an alternative. Anything to increase the length of time it takes to slow down.
While, one thing that might favour the ball, it could bounce many times, and each time it slows down its velocity across the surface of the Moon.
Maybe even combine all three technologies but not sure how.
Rob,
The problem with penatrators is that you’re now going to have to dig them out. Hitting at a steep angle with a good penetrator shape may require significant excavation equipment. I don’t see why this is going to be better than a shallow-angle impact with something more sphere shaped.
~Jon
Jon, okay but just for clarification, I was thinking about hitting at a shallow angle with penetrators.
Basically like Deep Space 2, but impact at a shallower angle and you can have a faster impact as with Deep Space 2 it was designed to impact Mars at 179 meters / second but the aim was for the equipment to survive and still be functional. Here we just need to have the metal intact and not vaporize so presumably could have a much faster impact. And deep impact would have buried itself 0.6 meters deep – not so difficult to dig out, this would be a shallow angle and you can design it sharp or blunt depending how deep you want it to go. It can go deeper and less damaged, or shallower and more damaged. Find the optimal depth so that it survives intact enough to be useful and not too deep below the surface. It’s just an idea.
https://en.wikipedia.org/wiki/Deep_Space_2
Paper on survival of small Cu projectiles hitting ice:
http://www.sciencedirect.com/science/article/pii/S001910351500603X
“[…] we study in the laboratory the survivability of 1 and 2 mm diameter copper projectiles fired onto ice at speeds between 1.00 and 7.05 km s−1. The projectile was recovered intact at speeds up to 1.50 km s−1, with no ductile deformation, but some surface pitting was observed. At 2.39 km s−1, the projectile showed increasing ductile deformation and broke into two parts. […]”
Paul,
Interesting find! If you ever want to do some blogging on SB, you’re invited!
Jon
What I get from that is that mechanical deformation, not thermal effects, is important at those speeds, so the density and strength of the target material matters (also the strength of the impactors). Regolith should be fluffed up to reduce peak stress.
a way of getting large chunks down that don’t require tons of smelting, refining, and alloy mixing infrastructure might be a good idea.
How did I miss that? Great point and thanks for clarifying.
What about a thick but light coating of aerogel? Would weigh very little and might cut not only peak deceleration but reduce the maximum impulse or shock the payload experiences.
Michael,
No reasonable sized coating covering a payload could make an appreciable difference at that velocity.
To put it in perspective, think of roughly a metre long “crumple zone” on your car at 110km/h (3+ feet at 55mph), to get the same proportionate effect at lunar impact velocity, you’d need a “crumple zone” roughly 50-100 kilometres across (~80mi). Of course, that’s human-survivable, but you get the idea.
I wonder if the regolith’s density could be transiently reduced, timed to coincide with the arrival of a payload. A simple way to do that might be to have a small penetrator (or several) arrive a fraction of a second beforehand, tossing a lower density cloud of regolith upward. The precursor(s) could be filled with explosives to increase the thrown mass.
Jon,
Something I meant to ask during your previous speculation about interacting with regolith to create magnetic drag for horizontal lunar landing: Instead of simple drag while using rockets to levitate, wouldn’t you be able to induce a magnetic levitating field if you had the appropriate arrangement of alternating electromagnets on the payload?
There’s a maglev scheme that uses this idea. With conventional maglev, you power an active magnetic field in the track, and then have the same (hence repulsive) field in the maglev cars. The alternative scheme instead has only passive elements in the track and a series of opposite magnetic fields in the cars, the speed of the maglev means that by the time a field is induced in one passive element in the track, the next (opposite) magnet on the car is passing over. Get the timing right and the net effect is repulsive.
[Of course it only works at speed, unlike conventional maglev, hence you still need wheels and engines to get up to maglev speed. So in terms of cost for Earth-based public transport, it’s apparently not much cheaper.]
Since the meteoric-iron grains act as independent elements, you might be able to achieve the same effect with raw regolith. Inducing a repulsive field in the regolith by using a series of spaced electromagnets on the bottom of the payload. (Reducing the spacing between the electromagnets as the payload decelerates.)
Eventually the payload is travelling too slowly for the effect to work (the minimum spacing, hence minimum possible speed, presumably depends on height above the conductive surface and the magnetic-density of the passive conductor), so you’d have to land on skids, or wheels if the surface is suitably prepared, or rocket levitation (and deceleration), for that last puny few hundred kilometres per hour of braking.
Once you’re are setting up major infrastructure, you could have a larger dedicated launch/landing sled with the magnetic system built in, but still much cheaper and easier than traditionally proposed mass-drivers. And with the possibility of having a system cheap enough to build, to be long enough, to be suitable for manned payloads.
So for launch: You accelerate the sled up to minimum maglev speed using some other system (doesn’t matter what, it’s only the first hundred kmh) to get up to maglev speed, then switch to a purpose-built passive maglev track for a few (dozen?) kilometres to get up to a few hundred kilometres per hour to get up to regolith-maglev speed. Then the sled leaves the built track and continues accelerating over a hundred kilometres or so of smoothed regolith to get up to orbital velocity, then a coasting section where the sled releases its payload. Then the whole system is mirrored for deceleration. (The sled decelerates for a few hundred kilometres over regolith, to get down to a few hundred kmh, then switches back to a built passive maglev track to get down to the minimum maglev speed, then switch back to an external deceleration system for the final less-than hundred km/h.)
For landing/capture, you do the same thing, but with the sled starting empty and “docking” with the income payload before decelerating.
By having the system completely mirrored, it works for both launch and capture. Although for capture, you’d want a longer coasting section to allow the incoming payload time to manoeuvre onto the sled at orbital velocity. Would a hundred kilometres of coasting be enough? Or would it be safer to launch the sled into a larger eccentric orbit and perform the capture at apolune, safely away from the surface, with the acceleration/deceleration sections back-to-back at perilune?
Of course, the density of grains of nickel-iron dust scattered in regolith might not be high enough density for any of this to work. And I have no idea how to work that out. Indeed, the passive elements might not be able to be grains, but might have to be looped-inductors. I’m not even sure of the formula to use to do a BOTE first approximation.
Paul D,
“The precursor(s) could be filled with explosives to increase the thrown mass.”
No advantage, impacts at lunar impact velocity release roughly the same energy as their own mass in TNT.
No advantage, impacts at lunar impact velocity release roughly the same energy as their own mass in TNT.
So making the precursor explosive might double the deposited energy. Seems worthwhile. Also, the explosive would produce gas, which I imagine would increase the amount of regolith excavated, particularly if it went off at depth.
“So making the precursor explosive might double the deposited energy.”
If it explodes before impact, there’s nothing to impact. If it doesn’t… there’s nothing left to explode.
“Also, the explosive would produce gas,”
So would nearly anything else.
If it explodes before impact, there’s nothing to impact. If it doesn’t… there’s nothing left to explode.
The energy in the explosive disappears into nothing in that second case? Has the law of conservation of energy been suspended?
So would nearly anything else.
The amount of gas produced just from the impact will be much lower for most materials.
“The energy in the explosive disappears into nothing in that second case?”
TNT releases 4MJ/kg of energy. C4 has 6. Wood has 16. Most fuel is in the 40-50 range (although you need to add oxidiser, which brings it down.)
Explosives don’t release more energy than the same mass of non-explosive CHON. Their party-trick is releasing it fast. But in a km/s impact, the speed of “release” is already taken care of: Chemical explosives are too slow.
“The amount of gas produced just from the impact will be much lower for most materials.”
Once vaporised, any plastic would produce the same amount of CHON volatiles as any plastique.
Pingback: The Slings and Arrows of Outrageous Lunar Transportation Schemes: Part 4–Propellantless Horizontal Soft-Landing Methods | Selenian Boondocks
“6. Where would you want material to be targeted relative to a settlement, to minimize risk to the settlement while simultaneously maximizing the odds of being able to recover the material easily?”
You could land material prior to establishing the settlement. Target the settlement region and shoot several material payloads to be recovered after soft landing by robots and astronauts.
Patrick,
Not a bad idea. Especially if you can transition to using the maglev landing approach after the colony is setup. But for getting initial resources down to a site, playing hypersonic space marbles with a few payloads may be a good way to get things started.
~Jon
The “artificial meteor” idea of landing stuff on a planet works even better on Mars since it has an atmosphere.
You could, for instance, launch tons of plastic chain to Mars. The chain unfurls before entry, and if it’s thin enough, it doesn’t melt and reaches a low enough terminal velocity that it lands gently on the surface so just needs a compressed-air cleaning to remove any dust. Then, just melt it down and you have nice plastic from Earth, perhaps something hard to synthesize like Ultem or polycarbonate.
Works for metals, too, of course.
The same idea would work with huge, thin hollow spheres, but that poses a packing challenge.
Paul D.: “I wonder if the regolith’s density could be transiently reduced, timed to coincide with the arrival of a payload. A simple way to do that might be to have a small penetrator (or several) arrive a fraction of a second beforehand, tossing a lower density cloud of regolith upward. The precursor(s) could be filled with explosives to increase the thrown mass.”
Make the precursor a dense (e.g., depleted uranium at 19 g/cm3) cylinder which impacts end-on. That will minimize the diameter of the resulting explosion, thereby lofting more regolith directly in front of the payload. It might be helpful for the front end of the precursor to be inverse-cone shaped, to concentrate the explosion in the manner of an armor-piercing shaped charge. The bottom of the payload might be coated with a sacrificial material chosen to maximize the gas produced, thereby creating a retro-rocket effect.
By the way, a resource that some may find useful in considering high-speed impacts is the on-line Crater Calculator developed by Keith Holsapple of the University of Washington:
keith.aa.washington.edu/craterdata/scaling/index.htm
Note that it handles both impacts and explosions: each impact can be mapped to a subsurface explosion, and vice versa. That tells you something right there about the nature of high-speed impacts!
I found an interesting material that could be used to make a low density “landing pad” for crashing payloads: foam glass.
This material is used industrially as a fairly strong insulating material. The stuff used here on Earth has density as low as 0.1 g/cc. It consists of a glass with small closed cells, and is produced by adding a foaming agent (sulfates, hydroxides) to gas melt. That density limit is due to the desire to maintain closed cells, but if that isn’t necessary I think the density could be made even lower.
On the moon, the amount of foaming agent needed could be even lower (due to arbitrarily low pressure). An area could be blanketed with low density glass foam simply by spraying a glass melt containing a dissolved foaming agent over the regolith. Ideally, the foaming agent should be something readily available, for example oxygen dissolved in the melt under pressure.
s/gas melt/glass melt/
It seems possible that the foam could be repaired between each landing with a single robot roughly resembling a paving machine. If it can be done with all local resources, it could considerably extend the timeline before needing more advanced softer techniques.
Many useful items can be packed to handle 1,000 gee loads. 1/4 second and 600 meters travel is not asking for a huge project if the foaming can be brought to a high TRL in advance of the early development.
I was still thinking of near-vertical impact of strong bulk materials (say, copper pellets) that would be used as inputs to lunar manufacturing. The purpose of the pad would be to prevent excessive fragmentation, not preserve large objects in an immediately usable state.
I see that foam glass has been proposed as a structural and insulating material for more general lunar construction.
One could also make foamed aluminum.
Here on Earth, hydrogen in aluminum has to be removed to prevent foaming when casting. This could be added on the moon, or (I think better) an aluminum-magnesium alloy could be used. The vapor pressure of magnesium isn’t that high, particularly if the mixture is an Al-Mg eutectic, but even a modest vapor pressure would cause foaming in vacuum. The Mg vapor in the foam’s cells would then freeze out on their inner surfaces.
Aluminium and magnesium are both available in large quantities on the Moon. This would make it an ISRU supplied process.