[ExI] Lunar dirt

[Eugen Leitl]

On Fri, Jan 07, 2011 at 11:25:25AM -0700, Keith Henson wrote:

> I understand the chemistry, perhaps better than most having made a
> number of metals and working a lot of chemistry in my misspent youth.
> Also have spent serious time inside monster processing plants like a
> 30,000 ton per day concentrator, copper and aluminum smelters, and oil
> refineries.  Plus a few power plants.

The question is one of scale. Bootstrap is about processing kg, not Gg.
Once you can process kg/day you can consider how you process 100 kg/day,
and after that 100 Mg/day. And only after then are you dealing with scales
comparable to our large terrestrial facilities.
 
> The question I have is how remotely run robots relate to a processing
> plant able to do something serious.  I.e., what are you proposing to
> *do* with them.

The initial stage is prospection. You build a modern version of the
Lunochod, deploy hundreds of these, and let them cruise the terrain.
Unlike Mars, turnaround for remote control is very quick. There's
plenty more of power (twice the insolation). Abrasion is higher, so
you have to expect shorter lifetime than Mars, unless you harden the
vehicles for lunar specifics (dust, electrostatics, radiation).

What interesting things could you do with small robots with 
centaur-like wheeled body plan, manipulators and nonphysical 
manipulators (electron and ebeam probes and manipulative probes)?

I think quite a lot. They would carry material to and fro the
central processing plant. They would prospect (mass spectroscopy,
electron and ion beam, solid-state laser), transport parts (sintered 
and cast), directly pattern substrate (sputtering, XY-electron and ion beam
forming) and such.

Think of them like termites or ants, and the central facility if
being the ant queen. The goal is for the swarm to grow until it can
produce a new clone on adjacent terrain. The eventual goal is
decentral control, emergent behaviour, in fact. The reason for that
is maximum behaviour complexity from minimal agent complexity.
 
> Although space elevators may not be possible from earth, they can be
> built out through L1 with Spectra, currently used for dental floss.

This is another point I wanted to make, but forgot. While space tethers
and hooks on Earth must be from unobtainium, and have to deal with the 
atmosphere commercial aramide is enough for the Moon. 

> One able to lift a thousand tons per day (using a moving cable design)
> can probably be constructed for a lower mass budget than a lunar seed.
>  It's still in the range of 100,000 tons, but it would lift it's own
> mass in 100 days.
> 
> Lunar elevator pretty much displaces magnetic launchers of all kinds.

The question is one of power, the idea is to saturate the Moon surface
with photovoltaics (alt.chrome.the.moon) and launchers, resulting in massive
enough launch facility so that the Moon starts losing mass.
 
> That's a big change from the days of Dr. O'Neill.
> 
> > From: Eugen Leitl <eugen at leitl.org>
> 
> > On Thu, Jan 06, 2011 at 06:21:14PM -0700, Keith Henson wrote:
> >> Will someone enlighten me about what remote manipulators on the moon
> >> are going to be doing?
> >
> > The first task would be exploration and mapping of the south
> > and north poles. The second part would be building large scale
> > thin-film PV arrays to mine volatiles and to build more
> > thin-film panels, and then to expand the industry base
> > until you can build linear motor launchers.
> 
> You can't build lunar mass drivers just anywhere.  Google achromatic

Right now we're still in early bootstrap. Early bootstrap is a lot 
different from late boostrap.

> orbits heppenheimer to see why.  So you need a road or something from
> the poles to the Lunar equator.  You also need a "catcher" which is a

You would start with rings around the poles (because only small terrain
features at the poles are semipermanently illuminated, so the amount
of power you harvest is limited), with the rings incrementally expanding 
towards the equator. 

The question is how much money you would be able to sink into the venture
until people start actually looking for ROI.

> massive structure in its own right.

You mean Earth-side? I don't see how, since you would launch powered
packets, with enough guiding logic and propulsion on board for aim
for aerobraking corridor, and then subsequent maneuvers. And, of course,
you can just use plasma thrusters once you've inserted into low lunar
orbit (with rocket corrections).
 
> >> You don't have a lot to work with; lunar dirt is about as far from
> >> useful objects as I can imagine.
> >
> > http://en.wikipedia.org/wiki/In-situ_resource_utilization
> 
> snip
> 
> That wasn't the question.  Specifically what are you doing with the
> robots to construct something useful>
> >
> >> I have followed this topic since the mid 1970 and, far as I know,
> >> there was never a believable flow chart with rock going in and useful
> >> stuff coming out the other.
> >
> > Keith, I thought your knowledge of chemistry and geology was
> > better than this.
> 
> snip
> 
> It good enough, I think, to call BS on vague handwaving.

The point is that there are hundreds of easily accessible papers which 
are anything but.
 
> >> Take solar cells.  Anyone have an idea of what sort of plant it takes
> >
> > I have a very good idea of that, yes. I've even toured the facilities.
> 
> The oil refinery like facilities where they purify the silicon?  Where
> did you find one that would give you a tour?  Or do you just mean the
> end stage where they mount cells?

No, it was a large facility (Wacker Burghausen) in mid 1980s. Of course
that one was on Earth, and it was silicon-specific. On the Moon, you would
heavily modify the parts or whole of the process. 
 
> >> make silicon?  What inputs the plant takes?  What has to be frequently
> >> replaced?
> >
> > Current CdTe takes about 10 g/m^2. That's 100 m^2/kg. 10^5 m^2/ton.
> > At 10% and 1.3 kW/m^2, that's 13 kW/kg, 1.3 MW/100 kg, or 13 MW/ton.
> >
> > Assuming you deliver 100 kg packages, each will be good for over
> > a MW of power.
> 
> What are you going to deposit the CdTe on?  How do you make it?  What

A dumb approach would be to melt sifted regolith in situ with a large
parabolic aluminized Mylar mirror, and sputter semiconductors on top.
A less dumb approach would be to use sifting, then remove iron from 
regolith with magnetic separation and electrostatic separation, then
reduce with hydrogen (electrolysis from polar cryotrap regolith), then
crushing and another magnetic separation step, then melting it into sheet
glass (why not float glass) or sinter plates on top of loose powder
and then lift off, then sputter. Add electron beam for processing
and ion beams for patterning.

The iron you could form into foil, which would be also a good substrate
for thin-film PV.

This *is* handwaving, but I have a hunch by running a production facility
for a year you will learn a lot more than in a decade worth of terra-side
lunar simulators.

> do you use for wires to get the power from where you make it to where

First-gen metals will be from Terra, second-gen will be aluminium from
kryolith-facilitated silica from regolith, iron, and whatever else you
can make. 


> you use it?
> 
> I am not saying it's impossible, just poorly thought out.  Few numbers

If I had a good plan I wouldn't be posting this to a public mailing list,
but doing a lot of knob polishing with the usual suspects. This isn't
my field, nor is this my project. Still, it is still heartening that
people in the mainstram are at all seriously pushing ISRU, even as
auxiliary/cost-saving methods for pure science and manned research
missions. Once this stuff flies I have a hunch this tail will start
wagging the dog.

> on power consumption, heat rejection, production rates, etc.  And
> don't forget the scaling problems up *or* down.

It is definitely a hard problem, and thankfully I'm not the one
who has to figure it all out. There are thousands out there who
eventually will. 

And we will watch, and marvel.

-- 
Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org
______________________________________________________________
ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE