> Mars is further away in travel time, but not much further in fuel costs, and it's far better than the Moon for living off the land.
The prospect of an inhabited Martian colony living off the land is dismal, for anything except extremely far-future timeframes (thousands of years or more). Any extraplanetary colony is going to be invariably tethered to Earth and highly dependent on imports for survival. The moon has the advantage of not just being closer, but having a launch window of "whenever you feel like it", in contrast to Martian launch windows that occur 26 months apart.
On the other hand, I'm curious how the delta-V numbers hold up for landing on the Moon, where aero-braking is impossible. Launch and landing require by far the most delta-V between either of the two bodies, much more than the difference between lunar transfer orbit and Martian transfer.
It's about 1.7km/s delta-v to land on the moon from LLO. If you are assuming an orbital station in the lunar gateway, that's another ~0.7km/s. From LEO that's another ~3.6km/s (3.2 to TLI and 0.4 to gateway). That gives you around 6km/s delta-v flat. Of course you can optimize this a bit by using a more direct route but it's probably cheaper if you leverage existing infra like the gateway.
Compare this to anything on mars. It'll be 3.6-4.0km/s to mars injection. Then 2.0km/s to low orbit (can be optimized with clever aerobreaking down to 1.0km/s). To the surface will be a minimum of another 0.4km/s (no engine, parachute only) and closer to 1km/s if you want a soft landing. So you can expect 5km/s at the bare minimum and ~7km/s along established "routes" leveraging to-be-existing infra.
So then the question becomes whether lower delta-v returns, regular launch opportunities, and shorter comms delay are more or less important than 1km/s delta-v savings.
I'd argue the delta-v savings aren't worth it on their own. Mars will eventually be worth exploring and it's a valuable opportunity but by every measure other than raw "throw it at the orbital body with no support or return plan" delta-v savings, the moon is a better site for initial extraterrestrial habitation studies.
>> It'll be 3.6-4.0km/s to mars injection. Then 2.0km/s to low orbit (can be optimized with clever aerobreaking down to 1.0km/s).
Except that all that delta at mars is done with aerobraking/aerocapture. Inbound probes can crash strait into the mars atmosphere without slowing down for orbital insertion. That's why it takes far less fuel to get to surface of mars than the moon, at least with objects car-sized or smaller. Only with larger/heavier objects is mar's atmosphere so thin that fuel must be used for capture prior to descent.
In the following deltaV map, much of the trip to mars is a free ride so long as you follow the red arrows. There are no free rides on the way to the moon.
For the record we have not done aero capture at Mars just yet - but aerobraking has been used a lot, after first capturing into a high eccentric orbit with an engine burn.
In any case, bodies with atmosphere do enable you to effectively shave up to half of the delta-v for a transfer when you know what you are doing. :)
I was just illustrating the path along the lunar gateway (LEO -> TLI -> Gateway -> LLO -> surface). Even if you do a direct (LEO -> TLI -> LLO -> surface) route, you still have to go through LLO as you'll need to be able to pick your landing spot and potentially even delay a landing (which is not possible if LLO is not budgeted in).
Actually now that I think about it, if you want to save delta-v, you can get an extra 25% off the TLI leg of the direct route with a low energy transfer using weak stability boundary trajectories. These savings however come at the cost of a significantly longer flight time and far slimmer margins for failure.
OTOH, we can build a lunar space elevator with existing materials because of the lower gravity (even despite the longer distance to L1). Not sure, can we manage that with Mars?
The physics works for Earth too. In all cases, you make the top of the cable bigger than the bottom to support all of the weight hanging from it. From what I've read, in the case of Earth and with current materials technology, we end up with the top of the cable having a diameter comparable to that of the Earth. Clearly, that's not feasible.
For Luna or Mars, gravity is reduced and the required diameter is less. Maybe it would even be feasible to build such an elevator if it were above the Earth. But now you're building above an alien plant, so you trade one set of potentially insurmountable obstacles for another.
I mean, that's a fair point. The total volume of the elevator cable would be greater than that of the Earth. The mass might still be less since we're not building it out of iron here, but effectively we'd have a binary planet with the centre of mass well outside the Earth's surface.
I'm not sure that that system would be unstable in human timeframes since the two would be tidally locked, although it would certainly alter the engineering stresses in ways that I'm grossly unqualified to calculate. I think a portion of the cable might be under compression rather than tension? I guess it depends on the rotational speed of the whole system.
Speaking of which, substantial amounts of energy would need to be spent accelerating the spin of the Earth/space elevator system to maintain a 24-hour day/night cycle.
However, Luna's presence would perturb the whole system, either tearing it apart with tidal stresses or being ejected from the system before that could happen.
I appreciate the correction. I'm not sure where I heard that particular piece of information, nor in that case what material was being examined. Perhaps that one was steel.
On planets, a space elevator goes to (geo)stationary[0] so that the cable doesn't wind up around the planet, but you can't do that on the Moon, because luna-stationary is occupied by the Earth, which inconveniently is too massive and spinning too fast to anchor the other side of the cable. However, the L1 point is also stationery relative to the lunar surface, and is the place where the gravity of the Earth and the Moon balance out.
[0] IIRC, geostationary specifically means Earth, but there's going to be some more general term for the same idea over generic parent objects and not just Earth
Lunar stationary orbit is around 88,400 km, which would be unstable for a satellite due to the Earth's gravity, but might allow for a space elevator pointed right at Earth to efficiently launch crates of helium-3 or hydroponic grain back to the planet.
Indeed. I've not even played with this in one of the many simulators, but I believe the suggestion is to put the counterweight a tiny bit closer to Earth than the L1 to stabilise it.
Although (and I wish I could find this again), I've read that lunar He3 is so diffuse that getting it out would incidentally give us so much purified aluminium, silicon, and oxygen, that sending all that back to Earth and magnetically decelerating it on arrival would give us more energy than the He3, as would burning those ingots with that oxygen.
Lunar stationary orbit is not 88,400 km, it's 384,000 km, i.e. the distance from the Moon to the Earth. The moon is tidally locked to the Earth.
Though you could equivalently try and go for Earth-Moon L1 with a counterweight on the other side of L1. It would be significantly more unstable though.
Even if I expected to be able to collect, I wouldn't give good odds on a bet that the moon or mars will physically still exist 1000 years from now.
Von Neumann machines have an existence proof in ourselves, and make it fairly fast to disassemble planets into a Dyson swarm. (I think Dyson swarms are a bad idea, and will grind to dust that blows away to interstellar space in the solar wind in geologically short timescales, but that's a different problem).
There is absolutely no reason to beleive that any kind of machine will ever exist that can "make it fairly fast" to disassemble an entire planet. You're probably entirely ignoring any kind of wear and tear on the machine itself, which could easily be outpaced by the rate of creating new machines for anything approaching the billions of years this would actually take using any imaginable fuel.
Assume humans are the VN machine. We can make space suits and greenhouses etc., so it's not crazy.
Assume reproduction such that population doubles every 25 years. Fast, but not insane.
(1000/25) = 40 doubling periods
8e9 * 2^40 * 100kg = 2.7 Mercury masses, or 12 lunar masses, or 1.3 Mars masses
That's a worst-case bound using a biological anchor. Dedicated VN machines can plausibly be faster: even if it's something weird like staying with biology and uplifting dogs, that's now 75 years, a plausible but currently hypothetical self-replicating 3D printer could make the timescales even shorter.
The sun provides enough power to do it in a week or two, though anything less than decades may have thermodynamics issues.
I'm not sure where your 8e9 number is coming from. Population doubling every year but the parents leaving the planet after mating?
In your calculation, after 1000 years, there would be 2^40 humans, each weighing 100kg - many orders of magnitude less than Mercury's mass.
Still, if we give it a few thousand more years of exponential growth, you will eventually reach such masses.
However, these numbers are meaningless - you are assuming that doubling the size of a population that's 2^39 individuals will take roughly the same time as a population that's 64 individuals individuals, which is not even close to plausible - especially when we reach ideas like a population weighing as much as half a planet.
You're also assuming that it's even possible for a mahcine to convert a significant proportion of a planet's crust to copies of itself - which is obviously false, as the crust is mostly rock, and machines require plenty of liquids and water to be produced (whether biological or mechanical or electronic). And that doing so will not affect the growth rate at all, even as the planet starts being formed of molten magma once all of the crust has been used up. And not to mention the gigantic earthquakes and supervolcaones they would have to deal with as a significant portion of continental mass gets shifted around.
Overall, you are only extrapolating some numbers to a completely absurd conclusion, and calling it plausible. There is nothing even close to realistic in your scenario, and indeed we have no idea if it's even close to possible to strip a planet down to create a Dyson swarm. I very much doubt your energy calculation as well, but that's already beside the point.
> I'm not sure where your 8e9 number is coming from. Population doubling every year but the parents leaving the planet after mating?
That's the current human population.
> In your calculation, after 1000 years, there would be 2^40 humans, each weighing 100kg - many orders of magnitude less than Mercury's mass.
You start with one human, I didn't. What are you even imagining that I'm describing, Adam-only parthenogenesis?
> Still, if we give it a few thousand more years of exponential growth, you will eventually reach such masses.
Even with Adam-only parthenogenesis, log2(8e9)*25 years is 822 years, less than one, definitely not plural, millennia.
Material science isn't my field, though it doesn't need to be given how many other places there are for whichever chemicals we want. Water? Oxides in the local rock, and four massive hydrogen gas giants (don't need much proportionally as H2O is 89% oxygen by mass).
> I very much doubt your energy calculation as well, but that's already beside the point.
Gravitational binding energy of Earth: 2.2e32 J; Mercury: 1.8e30 J; Mars: 4.9e30 J; Luna: 1.2e29 J
Luminosity Sol: 3.8e26 W
Time required to explosively disassemble (i.e. each part reaching escape velocity) each object: Earth: 6.6 days; Mercury: 1.3 hours; Mars: 3.6 hours; Luna: 313 seconds.
To preempt the obvious, yes I know that's the number for total luminosity and not the power available at any given moment given how many space habs have been built part way through the process, but that makes very little difference: Given the way the functions behave, you don't need most of the power of the sun until you can harness a significant percentage of it anyway.
I mean, this toy model also assumes that the only thing these humans do with their lives is reproduction, with the average individual adding only a little more than their own body mass to the VN swarm each generation, and not, e.g., building themselves a nice little space hab that's unlikely to mass less than 10,000 kg/person even if I make the grossly simplifying assumption of just adding life support to a tiny house or a camper van. It makes very little difference to exponential growth.
You haven't addressed the most important points at all, and keep coming up with toy models based on exponential growth. All of the models ignore the realities of how mechanical things work and how they can break and how they actually operate (just for a basic example, there is no mechanical system we have any idea how to build that could move any amount of the Earth's mantle in any way, since the mechanisms would simply melt) and rely on the ridiculous idea that this exponential growth can actually be maintained indefinitely to paper over various other omissions.
You also don't need to be a materials scientist to know that you can't get water or oxygen out of rock with sheer mechanical force.
Your estimate for the energy of the sun takes into account all of the energy sent in all directions in all spectra. The amount reaching the earth is significantly less - 1.73e15 W, or about 10^9 times less - and the amount that can realistically be captured is far less than that.
Overall, don't worry: there is exactly 0 chance that any human advancement will disassemble even a dwarf planet in the next millennium in the real world. Just because Freeman Dyson could write some back of the napkin computations it doesn't mean this is actually possible in any meaningful way.
> just for a basic example, there is no mechanical system we have any idea how to build that could move any amount of the Earth's mantle in any way, since the mechanisms would simply melt
You know stuff cools down, right? Power loss to radiation is proportional to T^4.
> You also don't need to be a materials scientist to know that you can't get water or oxygen out of rock with sheer mechanical force.
Good thing you're putting words into my mouth, then. Hint 1: How do we do this for aluminium? Hint 2: I didn't say "mechanical" for this.
> Your estimate for the energy of the sun takes into account all of the energy sent in all directions in all spectra
I know, and I said as much with different words.
Do you perchance know what a mirror is? Or how light they are? How little of (insert-planet-here)'s mass you need to turn into PV and/or mirrors to get to covering the planet, how little time it takes to use those to gather the energy needed to run a launch loop to get a second planet-tiling-quantity to orbit?
That's why I preemptively made the point that you're ignoring here.
> You know stuff cools down, right? Power loss to radiation is proportional to T^4.
It only took around 160 million years for the Earth's crust to form, so yeah, sure, stuff cools down, eventually.
> Hint 1: How do we do this for aluminium? Hint 2: I didn't say "mechanical" for this.
Ok, mechanical was my idea - but chemical extraction of oxygen requires some other compounds to form, potentially making the whole thing even less usable for future conversion into more copies. Plus, it requires an input of some other materials, which may not be easy to create.
> Do you perchance know what a mirror is? Or how light they are? How little of (insert-planet-here)'s mass you need to turn into PV and/or mirrors to get to covering the planet, how little time it takes to use those to gather the energy needed to run a launch loop to get a second planet-tiling-quantity to orbit?
That still only gives you the 10^15 watts that reach the Earth, not the 10^26 number you were citing. Also, covering the whole planet with mirrors or PVs is again not nearly as trivial as you make it out to be, and this "launch loop" idea is just some abstract design, not something we can actually build (despite what the author would have you believe).
> Assume reproduction such that population doubles every 25 years. Fast, but not insane.
This is already completely implausible given everything we know about human behavior, but it reaches impossibility very quickly when you consider the possibility of humans becoming more than a negligible fraction of the mass of their single host planet. We aren't machines that can trivially reproduce ourselves from commonly available materials and then eject into space. Feeding ourselves is hard, getting to space (alive) is harder. And once the overwhelming majority of us are in space because there's no more room down below, how are we supposed to meet up to keep up the 25 year doubling rate. How are we supposed to keep up the rate of resource extraction from Earth?
> We aren't machines that can trivially reproduce ourselves from commonly available materials and then eject into space.
Nah, we use plants to turn raw materials into what we can consume. And in the other direction, we can only make stuff on this scale with factories that take a while to build. But in both cases, that's a distinction without a difference. A farm and a factory rather than a spacesuit, makes no difference on this scale, so long as they feed themselves while growing their families.
> Feeding ourselves is hard, getting to space (alive) is harder.
Feeding ourselves is about 1% of our current labour. Getting into space is only hard because we use rockets, but at this scale we'd use launch loops, atlas towers, orbital rings, or similar. Those are extremely cheap, like "$300 to LEO" cheap for this thought experiment's ("spherical cow in a vacuum" model of a) 100 kg human.
> And once the overwhelming majority of us are in space because there's no more room down below, how are we supposed to meet up to keep up the 25 year doubling rate. How are we supposed to keep up the rate of resource extraction from Earth?
There's lots of ways I've seen suggested. Even without the exotic options like the Dyson Motor (would take too long, at 40k years for Earth, not seen the numbers for Luna or Mars) or redirecting Kupier Belt Objects to blow off percentage points of the target planet mass at a time, even just with traditional digging, at that scale it's "how fast can you drill vertically?" and "how many launch loops can you wrap the target planet in?", followed by "how fast does the deep ground cool down when exposed?" — the latter being why I said thermodynamics probably gets in the way when the timeline gets down to decades; this is radiative-dominated cooling in a better vacuum (insulator!) than most laboratories let alone thermos-flasks.
Interestingly, rocky planet disassembly isn't horribly complicated and is a well solved problem. Just need large scale rail and satellite manufacturing.
Only requires more structures and geoengineering larger than humans have bever done before.
We know how, and we know it doesn't require more than operating at higher scales of industry than ever before. Any interplanetary or kardishev 1+ venture makes those assumptions.
The prospect of an inhabited Martian colony living off the land is dismal, for anything except extremely far-future timeframes (thousands of years or more). Any extraplanetary colony is going to be invariably tethered to Earth and highly dependent on imports for survival. The moon has the advantage of not just being closer, but having a launch window of "whenever you feel like it", in contrast to Martian launch windows that occur 26 months apart.