The paper's author Dr. Phil Metzger is such a rockstar. He is The Expert on the mechanics of soil erosion by rocket exhausts, and writes a lot about the problem in an approachable way, https://twitter.com/DrPhiltill/status/1658507854859337737
It turns out that the rate & mechanics of erosion by rocket plumes is an unsolved problem that requires a new kind of model. To quote from his thread,
In the Apollo era the thinking was that the rate of soil erosion is controlled by conservation of momentum. It turns out this is wrong.
NASA researcher Leonard Roberts, the first person to research this topic, hypothesized that the soil grains steal momentum from the gas, which slows down the gas and thus reduces the erosion rate. It was this feedback that determined the rate.
I argued some years ago this has to be wrong because the particles achieve their high velocities far downstream of where they are lifted off the surface, so momentum transfer does not provide feedback to control the rate that grains are lifted.
He's going to be publishing his alternate model soon-ish. Can't wait to see what he has come up with.
"...This manuscript analyzes lunar lander soil erosion models and trajectory models to calculate how much damage will occur to spacecraft orbiting in the vicinity of the Moon. The soil erosion models have considerable uncertainty due to gaps in our understanding of the basic physics. The results for ~40 t landers show that the Lunar Orbital Gateway will be impacted by 1000s to 10,000s of particles per square meter but the particle sizes are very small and the impact velocity is low so the damage will be slight. However, a spacecraft in Low Lunar Orbit that happens to pass through the ejecta sheet will sustain extensive damage with hundreds of millions of impacts per square meter..."
Well lunar gravity varies a huge amount by location, it may be possible for some of it to somehow end up in actual orbits, especially when it's intentionally done so.
My understanding was that those variations in gravity ("mascons") rendered stable lunar orbits next to impossible. On Apollo missions, orbits degraded by many kilometers over the course of just a few days.
Interestingly, a researcher looking for probable impact sites found Apollo 11's ascent module could potentially still be in orbit. (Though, it's probably not.)
The mechanism would loft a cloud of particles with a wide range of trajectories through a notoriously uneven gravitational field.
The odds that a sufficient fraction of particles would remain in orbit for long enough to make routine operations in low lunar orbit or on the surface itself seems ... plausible.
And of course, the situation could be compounded by multiple burns and/or at multiple points on the surface, at a somewhat increased cost to the attacker.
Actually, Moon is rather immune to Kessler syndrone due to its bumpy gravity caused by sub-surface mass concentrations.
So Kessler syndrome could certainly develop for a whiley but would be cleaned up rather quickly as all the pulling & pushing of the "rough" gravity converts the orbital speed into heat, until all the fragments impact the surface.
Might be a bit more dangerous on the surface for a while though, with lot of stuff striking it at near orbital speed in an almost horizontal direction. That could ruin your evening stroll quite badly.
IIRC single digit years probably - Apollo missions released a couple sub satellites and missions control was then very surprised when those satellites lost altitude and crashed in a matter of months.
In comparison, there is likely still stuff from the 60s in orbit around Mars, and that's for a body with (thin) atmosphere.
Kessler is a chain reaction. Destruction is caused by secondary effects, i.e. bits of satellites the primary projectile broke hitting targets. This is closer to an area denial weapon: the destruction is caused directly by the debris blown off the surface.
It is also not permanent. The orbits of the debris would intersect the point where they depart, which is pretty close to the engine. Basically the engine would be hitting itself with everything it fired.
Therefore, the logical thing to do is to put it on the correct side of a mountain, to shield the engine. But that would also collect all the debris. So it would generally only be in orbit for one orbit.
(To be honest, I think on any planetary body without an atmosphere, long term everyone is going to have to dig in to the planet, and to a non-trivial degree, too, not least of which is the complete indefensibility of surface installations.)
The expansion of the gas after it leaves the nozzle in vacuum would give the particles an additional kick. I'm not sure if their orbit would still intersect the engine or effectively boost higher.
I left myself some wiggle room in the phrase "where they depart" for that reason. It won't all be a straight line out of the rocket motor or whatever is pushing, because in the first fractions of a second the gasses and the particles can interact and bash each other into slightly different orbits.
However, that will dissipate quickly and you'll certainly be looking at a set of orbits that all pass through something relatively close to the origin. They're not going to be interacting for the first time a quarter of the way through the orbit and bouncing around a lot there.
No, it has an exosphere. The difference being an exosphere gas particles are more likely to collide with the ground at the end of a parabolic trajectory than to collide with other gas molecules, so you don't really get any of the properties characteristic of atmospheres.
This problem has been known for a long while. The models used here by Metzger have such a large uncertainty and only take into account erosion due to shear. As soon as the erosion transitions from shear-based (smooth sheets) to bulk (fluidized), none of the data extends to that regime. A massive vehicle landing on the moon will definitely cause fluidization. To estimate erosion and ejecta needs far more detailed numerical methods [0].
Small and medium meteorites smack into the moon's surface all the time, being it has no atmosphere. I find it hard to believe that human-built landers have nearly as much impact on low-orbit grit than these meteor impacts.
Impacts are impacts. Rockets are more like leaf blowers. Drop a massive rock into a pile of leaves and few leaves even move. Point a leaf blower at the pile and leaves will scatter everywhere.
Some of the data from LADEE and Lunar Dust Experiment (LDEX) instrument might point to this being the case, but it was more along the lines of human-built landers do not have nearly as much of an impact as meteor showers themselves (not necessarily the impact and resulting dust).
"if LADEE did encounter any lunar soil particles thrown up by the final descent of Chang'e 3, they would have been lost in the background of Geminid-produced events." [0]
That said, the Chang'e 3 is an order of magnitude (or close to two) smaller than the lunar landers they are talking about in the study. Also my own speculation is that the more continuous thrust of a lander may get particles to higher velocities due to the additional time for acceleration in the wake of the thrust as compared to the single impact of the meteor.
I struggle to compare exactly how bad the lunar dust ejection is though. Most Micrometeoroid and Orbital Debris (MMOD) curves are specified as a Flux by particle size (velocity is sort of irrelevant, as you assume most of the velocity is from the spcecraft itself and most hits are in the direction of travel of spacecraft, the ram direction). My suspicion is that MMOD flux in a LEO orbit is still going to be far far worse.
Edit: The paper talks about flux of particles 10 um and smaller of about 10,000 impacts/m^2 during the passes. If we assume that this is a sphere of iron (new MMOD fluxes are specified in mass, not size) its ~5e-9g. In LEO at 400 km altitude (a little above the ISS) the flux of particle this size is ~1000 impacts/m^2/year. But the paper says smaller than <10 um. And at smaller masses the flux increases exponentially to 10^7 particles/m^2/year at a particle mass of 10^-18 g. So I believe my suspicion is correct that most LEO orbits are still worse, but its hard to compare apples to apples.
Landers don't impact hard - they fire rocket engines down, which may be much more efficient at kicking up dust than an impact.
(I still think this is overblowing the problem, because any lander that causes this big of an ejecta problem would also badly damage itself. All the designs will put a LOT of engineering work into minimizing debris, eg Starship putting separate landing engines high up on the vehicle.)
> because any lander that causes this big of an ejecta problem would also badly damage itself
Not necessarily because the relative speeds will be very slow. Not so in low lunar orbit, where an orbiting spacecraft will slam into the ejecta curtain at >1 km/s.
The paper estimates that if there's LLO debris, it'll be starting out with about 1.6km/s of surface-relative velocity. Not something you want to get even a small percentage of on your landing gear.
The ejecta would go out to the side and not harm the lander. I'm just amazed that the ejecta is being thrown hard enough to be a threat at orbital altitude.
Velocity required to reach a 50 km orbit* above the moon's surface is only ~1.6 km/sec, and there's no air resistance to slow dust particles kicked up by the craft.
For Mars, the orbital velocity is ~3.5 km/sec, thus requiring almost 5x the energy for a given mass of detritus (E = 1/2mv^2); and while its atmosphere isn't as thick as Earth's, it'll definitely cause drag for particles going that fast.
* You don't quite need orbital velocity for a plume to get high enough to disrupt an orbiting craft, but it's a handy reference point.
The lander size under consideration is about 40x that of the Martian rover, I don’t know that a sky crane would work as well without a parachuting stage and atmosphere, and finally it seems rather unhelpful for taking off.
Afaik that's mainly so the plume of razor sharp dust doesn't tear their engines to billion tiny pieces. Unlike the LEM, they won't be bringing a spare for liftoff.
I wonder how much Starship mitigates that by having the descent propulsion controlled by the top thrusters. I'd assume they use the same thrusters from take off too as we've seen the damage Raptor 2's can do, let alone 3 and whatever comes after the 350bar line.
Also, could regolith be bound or, well, packed down to build pads?
There is a project to inject small aluminum granules into exhaust, so they melt when leaving engine but solidify on contact, which would be a good way to cover your landing site with solid aluminum layer. A rocket engine could easily melt a ton of aluminum per minute.
If you hover high enough in vacuum, gases have enough time to disperse and won't throw ejecta that much, but those melted aluminum droplets can travel relatively unimpended and solidify on contact. You land only when you have good solid surface.
Spallation is a process in which fragments of material (spall) are ejected from a body due to impact or stress. In the context of impact mechanics it describes ejection of material from a target during impact by a projectile
Sure, the article doesn't say that something would be hanging around the moon permanently, it evaluates the damage that might occur if the orbiting spacecraft passes through the splash of debris ('the ejecta sheet') as it's happening - it doesn't really attempt to evaluate how likely it is to hit it, the discussion is about the expected consequences if it happens.
Although it does assert that the NASA Gateway orbit passing through the ejecta sheet "will probably be several times before the sheet is dispersed."
Yes, that's why I said "only in the immediate aftermath of a landing".
The problem is, your orbiter is necessarily in orbit at the same time that your lander is making its landing. And the lander kicks up debris that can then threaten the orbiter.
Yes, I know about the masscons, but it seems highly unlikely they'll be just at the right point. It's hard to believe that any object given a single impulse at ground level, in a vacuum, is not going to hit the moon again in one orbit (unless it is given escape velocity).
Masscons did perturb the Apollo missions enough that they had to switch to a doppler radio to navigate accurately.
The paper models a 40 t lander. That's roughly Blue Moon. The sum of all Chinese probes landed on the moon is far less than that, and I wouldn't be surprised if the sum of all Soviet Luna landers was also less than that.
As noted, the paper is referring to a 40 t lander with a single 67 kN thrust engine.
As some other data points, the Apollo landers were 16t (later ones a bit more for an extended mission). The ascent module had a dry mass of 2445 kg and had an additional 2376 kg of propellant (5t).
The lander's descent propulsion system was capable of 10,500 lbf (47 kN) that could be throttled between 10% - 60% and 100%. (1,050 lbf (4.7 kN) and 6,825 lbf (30.36 kN))
The ascent propulsion system was 3,500 lbf (16 kN).
It turns out that the rate & mechanics of erosion by rocket plumes is an unsolved problem that requires a new kind of model. To quote from his thread,
He's going to be publishing his alternate model soon-ish. Can't wait to see what he has come up with.