PDA

View Full Version : Industrialization of Iapetus, moon of Saturn



IsaacKuo
2015-Jan-20, 12:01 AM
Industrialization of Iapetus, moon of Saturn

Iapetus is an interesting target for industrialization of space. Much of its surface is pure water ice, and it's cold enough for water ice to be a robust construction material. Furthermore, ice is optically clear enough to form efficient solar concentrator fresnel lenses. The slow rotation period of 80 days mollifies the issue of Sun tracking. The very low gravity and lack of wind mean that ice walls can form extremely tall fresnel lens arrays.

The basic lens element is a narrow prism, produced by freezing water in a tilted rectangular tray. The amount of tilt determines the angle of the prism, so a single tray is a suitable mold for all fresnel lens sections. Two layers of prisms--one aligned horizontally, and one aligned vertically--provides sunlight concentration in two dimensions. As more and more sections are added, more and more solar heat is generated, increasing the rate at which liquid water is generated. This, in turn, increases the rate at which more solar power concentrators are created. This positive feedback loop provides rapid exponential scaling. Even though sunlight is only 1% as powerful out there, this exponential growth can soon produce 100 times the area of solar array you could have launched from Earth.

The Sun very slowly creeps across the sky; the focal point of the sunlight very slowly sweeps across the ground. The mining robot periodically shifts position, removing its "melt straw" from its current position, and inserting it into a new position to slurp up water from a new melt cave. The robot basically runs on steam power, with exhaust steam used to melt more water from the melt cave, and the slurped up water used to make more fresnel lens walls.

The steam power is used to helix melt large cylindrical slugs out of the ground--leaving behind large cylindrical tunnels suitable for use as a space gun. Hydrogen/oxygen gas is generated via electrolysis, and placed in the closed end of the tunnel. Igniting the hydrogen/oxygen mix pushes the ice slug out of the tunnel at around 1.5km/s (a near escape transfer orbit).

From there, a solar thermal steam thruster provides sufficient performance to bring the ice ship to Jupiter--only about 0.04km/s needed. At Jupiter, a free gravity assist boosts the ship toward either Earth or Venus--aerocapture and/or a series of Earth/Venus gravity assists slow down the ice ship into Earth or Venus orbit. Basically, do what the Galileo probe did in reverse, doing an Earth-Earth-Venus series of flyby maneuvers to slow down the ice ship (assuming direct aerocapture is not considered acceptable).

Compared to potential inner system sources, Iapetus has far better rapid scalability potential. Even if they had pure ice at the surface, sunlight is too strong for ice to be a robust construction material of solar concentrators in the inner solar system. Compared to other icy worlds of the outer solar system, Iapetus has the lowest delta-v costs by far. Iapetus is also almost unique in its low rotation rate, making fixed solar concentrators practical.

No complex chemical reactions or rock crunching is required. Just some steam/water pipes, solar concentrated heat, and melting stuff. And a couple electrodes for simple electrolysis of hydrogen/oxygen gas (no need for compression, cryogenic cooling, or fancy insulation...just a spark to ignite it). It's a way to get large amounts of ice to Earth orbit, via locally produced ice ships blasted into space from a huge ice tunnel...

cjameshuff
2015-Jan-20, 02:11 AM
I'd expect lots of problems with stray heat/vapor causing problematic sublimation/redeposition. This does bring up something I've thought about on occasion...a cold water vapor engine using sublimation of ice. Abysmal Carnot efficiency, of course, but perhaps useful for sun tracking in microgravity.

Strips or bumps of ice on photovoltaic panels could focus the light with much lower losses and more tolerance of imperfect shape and surface damage. Just import a thin film substrate that can be layered on pykrete backing and have the lenslet array deposited on the front.

IsaacKuo
2015-Jan-20, 05:37 PM
I'd expect lots of problems with stray heat/vapor causing problematic sublimation/redeposition.

Where is this stray heat supposed to come from? The mining robots would be operating in a vacuum, so everything is naturally surrounded by an insulating vacuum. The main potential weak point, I think, would be the insulation around the mining pipe. You wouldn't want the pipe to melt out a hole around it (spoiling the seal). Instead, you'd want a coaxial pipe with a vacuum insulation barrier.

Still...I do imagine open venting of steam to be a part of the process of producing the ice prism slabs. Each slab is produced withing a slightly conical tube. The tube includes a couple mold inserts--each occupies almost half of the tube, leaving a slab-shaped gap between them. Tilting the molds with respect to each other provides the prism angle. I see the mold inserts as being hollow so they can draw heat away from the slab via evaporative cooling. Repeatedly fill the hollow with some water, and then vent it to space to evaporate it.

This is going to spray steam/snow out the exhaust vents. Obviously, you'll want to take care to point the exhausts at an angle so it doesn't cause a problem.


This does bring up something I've thought about on occasion...a cold water vapor engine using sublimation of ice. Abysmal Carnot efficiency, of course, but perhaps useful for sun tracking in microgravity.

I'm not sure what you're thinking of, here, or why it necessarily would have poor efficiency.


Strips or bumps of ice on photovoltaic panels could focus the light with much lower losses and more tolerance of imperfect shape and surface damage. Just import a thin film substrate that can be layered on pykrete backing and have the lenslet array deposited on the front.

Ice bumps won't work, since they'll be melted by the PV panels at the focus. You could separate them by vacuum, though. Still, I don't really see the benefit. The losses are pretty much the same whether the focal distance is 1cm or 100m. Having a large scale fresnel lens, though, lets the target be much larger and it's more tolerant of imperfections. Larger scale is important for the desired application of melting ice.

IsaacKuo
2015-Jan-20, 07:50 PM
I just realized that it's much simpler to extrude the ice prisms than to mold them all at once. The extruder is a short section of pipe with two mold inserts to form the slab sides at the desired prism angle (only one of the inserts needs to be adjustable). Liquid water enters one end; its pressure forces the ice slab out the other end.

The mold inserts use evaporative cooling to draw heat from the slab, freezing the new water as the slab is extruded.

With this extrusion process, very long prism sections may be produced. These are laid out in two layers perpendicular to each other to create large square fresnel lens panels.

cjameshuff
2015-Jan-20, 11:02 PM
Where is this stray heat supposed to come from? The mining robots would be operating in a vacuum, so everything is naturally surrounded by an insulating vacuum. The main potential weak point, I think, would be the insulation around the mining pipe. You wouldn't want the pipe to melt out a hole around it (spoiling the seal). Instead, you'd want a coaxial pipe with a vacuum insulation barrier.

You're talking about focusing light to generate steam. There's all sorts of places for heat to get out where you don't want it. Just diffuse radiation coming from the concentrator focus could damage your optical surfaces directly or through causing sublimation which then produces frost.



This is going to spray steam/snow out the exhaust vents. Obviously, you'll want to take care to point the exhausts at an angle so it doesn't cause a problem.

In vacuum, you're going to need to do more than point the exhausts at an angle.



I'm not sure what you're thinking of, here, or why it necessarily would have poor efficiency.

A low pressure engine driven by sublimation of ice would inherently have a low hot-side temperature. There isn't much room for a large temperature delta. The low pressures would also mean relatively large friction losses.



Ice bumps won't work, since they'll be melted by the PV panels at the focus. You could separate them by vacuum, though. Still, I don't really see the benefit. The losses are pretty much the same whether the focal distance is 1cm or 100m. Having a large scale fresnel lens, though, lets the target be much larger and it's more tolerant of imperfections. Larger scale is important for the desired application of melting ice.

There's no reason they'd melt. The average heating over the surface of the panel would be low, heat would be produced at small points within the ice and could be dissipated both through the front and back, and the concentration doesn't need to be high enough to produce steam at the focus. The main benefit is greater simplicity and tolerance of imperfections...the large scale fresnel lens array certainly does not have the advantage there...and less trouble with insulation, thermal radiation, high pressure steam, etc.

IsaacKuo
2015-Jan-21, 06:09 PM
You're talking about focusing light to generate steam. There's all sorts of places for heat to get out where you don't want it. Just diffuse radiation coming from the concentrator focus could damage your optical surfaces directly or through causing sublimation which then produces frost.

That sounds impossible. Even if the light absorbing heating chamber were 100% reflective, it would only emit as much light as originally hit it. This would at most double the amount of light passing through the ice prisms. With sunlight only 1% as strong as sunlight here on Earth, this is nowhere near bright enough to bring up equilibrium temperature to a problem.

But of course it makes no sense for the light absorbing heating chamber to be 100% reflective. Instead, it would use a target which is extremely black, like carbon soot. The amount of EM radiation emitted will basically depend upon the desired operating temperature. Hot steam is desirable if it's meant to power a steam turbine, but warm water is good for producing ice prisms (minimized radiative heat losses from the pipes).


In vacuum, you're going to need to do more than point the exhausts at an angle.

Why? All that really matters is that its pointed away from the concentrator wall, some dozens of meters away. If any frost gets on the outer surface of the probe, that's no problem. The only place where it would have any effect on its functioning would be on the sunlight heating target--but it's warm enough to sublimate any frost that lands on it in moments.


A low pressure engine driven by sublimation of ice would inherently have a low hot-side temperature. There isn't much room for a large temperature delta. The low pressures would also mean relatively large friction losses.

I think I get it now. I don't really understand the point of it, compared to simply continuing to increase the pressure and temperature. But I suppose it could be used where you have a source of low grade heat.


There's no reason they'd melt. The average heating over the surface of the panel would be low, heat would be produced at small points within the ice and could be dissipated both through the front and back, and the concentration doesn't need to be high enough to produce steam at the focus.

You're talking about direct contact between the PV panel and the ice, so the ice at contact will melt unless the PV bit is kept cool enough to prevent it. However, you recommended backing it with pykrete, which is an insulator. This means the PV bit is surrounded on all sides by insulator that is opaque to the relevant infrared thermal wavelengths that might have allowed for radiative heat loss.


The main benefit is greater simplicity and tolerance of imperfections...the large scale fresnel lens array certainly does not have the advantage there...and less trouble with insulation, thermal radiation, high pressure steam, etc.

The slightest imperfection in the bump, and the concentrated sunlight will miss the PV bit. There are two basic strategies you could use; you don't specify which one. If we assume that wires are cheap (locally produced?), then you could dot the PV bits in a 2D grid of dots--wired together with lots of wires. This means that each PV bit might be, say, 1mm by 1mm, behind a 1cm bump. This produces a concentration equivalent of sunlight here on Earth, with pretty easy aiming requirements. Unfortunately, a 1cm thickness is too insulative to avoid melting, but the design can be adjusted to place a vacuum gap between the PV bits and the ice lenses.

But this scheme assumes that the wires between the PV bits are cheap. Otherwise, it makes sense to concentrate all the bits in one place, and use a larger fresnel lens (like I suggested in the first place).

The other basic strategy is 1 dimensional light focusing. In this case, you could have a 0.1mm wide PV strip behind a 1cm wide bump. Again, this only produces the same concentration of sunlight here on Earth, and again the bump thickness is too insulative to avoid melting. But now the width of the PV strip is so small that the slight imperfections and bad aim will cause the sunlight to entirely miss the strip.

With the fresnel prism wall I envision, each prism is perhaps 10cm wide, with a heating target that's about 10cm by 10cm. Small imperfections just mean a reduced amount of sunlight hits the target--but you can compensate for it simply by expanding the wall and/or having a larger target. Each wall section might be 10m tall by 1m wide. This provides a concentration of sunlight which is 10x the light here on Earth. By placing ten of these side-by-side, you get 100x the concentration of sunlight here on Earth. That's a massive improvement in performance compared to the bump idea, and it produces all of its power in one place--no massive grids of wires required.

The reason this scales so well is because the Sun's rays are nearly parallel at this distance. This limits beam spread, meaning the prism wall can be pretty far from the target even if the target is pretty small. For a target 10cm wide, the wall can be 100m away. This means you can scale up a LOT of wall sections side-by-side. In principle, you could have around 100 of these sections side-by-side, providing the equivalent of 1000x the concentration of sunlight here on Earth. You only really need a tiny fraction of this, so you can overwhelm any imperfections with brute force addition. Realistically, the target will likely be more like 20cm by 20cm in size. The system will be functional even with only 1x Earth sunlight (sufficient to heat up hot water, which is then used to melt more ice and create more ice prisms).

Your idea of bumps on a PV array does not offer the potential to scale up like this, or to overcome imperfections with brute force. In order to gather sunlight from a 10m by 100m sized wall, you need enough PV bits/wires to cover that entire area. This system only needs a compact mining probe with a target that's on the order of 10cm in size.