Talk:NIAC: Aerospace products
From Wise Nano
Come up with a cooler, more succinct name.
Creating Just-In-Time Parts From Generic Nanoscale Modules
--Selenite 15:41, 28 Sep 2004 (CDT)
Thanks, that name sounds good to me. by Chris Phoenix, CRN 16:18, 28 Sep 2004 (CDT)
And thanks for being the first person (other than the investigators) to post on this project.
Glad to help. I'm tackling TT-F's report next, hopefully I'll be able to give more help down the line. --Selenite 16:31, 28 Sep 2004 (CDT)
(no topic) by Chris Phoenix, CRN 17:59, 28 Sep 2004 (CDT)
Karl, I've fixed the bug that kept your post's headlines from showing up; now if you don't type a subject, it'll fill in as you see here. And your signature will be automatically added to the header.
Thanks! by Selenite 16:52, 29 Sep 2004 (CDT)
Actually, that was user error, I'm not used to adding a subject for comments.
Self-replication time by Tom Craver 21:18, 1 Oct 2004 (CDT)
Does the "couple hours" self-replication time take into account the idea of replication from pre-made nanoscale modules? If a nanopart contains roughly 100K atoms (100x100x100 but with significant empty space), that's only 10000 nanoparts to assemble into a 1B atom assembler robot. At that scale, it seems like this might be accomplished in seconds - or less?
Since heat disposal can be a major issue in space, continuous production of nano modules at a low heat dissipation rate (to enable rapid construction of products as needed) seems like an excellent strategy, worth mentioning in the context of manufacturing using raw materials found in space. It's probably also worth covering the applicability of the approach to materials commonly found on the moon or rocky asteroids - beyond carbon. I believe Brett Bellmore mentioned sapphire. Also, NASA will probably be interested in applicability to more conventional materials, such as metals.
Do you have material strength estimates for components made up of nano cubes that attach at their faces? Seems like ~10% of a surface area might contribute strength to attachment - does that imply 1/10th the tensile strength of the base material?
A key "proof point" - if you could create a rocket nozzle using nano modules, you'd have a very convincing point. The extreme thermal, reactive gas flow, and pressure conditions make me question the feasibility. Failure on this point will not kill the concept, but success would make it a "killer concept" - there are likely few other components that combine such high mission criticality with such high stress conditions. A fall-back killer application would be spacesuit or habitat production/repair - less stressed, but equally critical - but I suspect NASA will want to know if nano modules cannot be used to make rockets. (I certainly do.)
Nano-module rocket nozzle by Brett Bellmore 09:54, 2 Oct 2004 (CDT)
The chief design challenge for rocket nozzles is incredibly high heat loading. There are essentially 3 approaches used in nozzle design:
1. Massive heat sink. Build the nozzle out of something with a high thermal conductivity, and considerable thermal mass. As the rocket fires, the nozzle gradually soaks up heat, rising in temperature until the inner lining exceeds some critical temperature, and starts degrading. This works very well indeed for rockets which fire for short periods, such as reaction control rockets.
2. Ablative materials: The suface of the nozzle degrades, consuming heat in endothermic reactions, and sheilding itself with a sheath of reaction products. This is the approach generally used in high performance single use rocket engines. It has the downside that it only works so long, and nozzle geometry changes during the engine firing. Think the Space Shuttle solid fuel boosters.
3. Active cooling: A very thin, high thermal conductivity lining, is exposed to the rocket exaust on one side, and to the (Cryogenic?) fuel or oxidizer on the other side. The upside of this, is that if it works, the engine quickly achieves thermal equilibrium, and can fire indefinately. The downside is that the liner is subject to incredible thermal stresses, and will frequently be good for only so many firing cycles, before accumulated stress degrades it's mechanical properties too much. This is currently used in high performance liquid fueled engines, such as the Space Shuttle main engines.
A fourth approach, which OUGHT to be superior to the above, and enable an engine to survive any mumber of firing cycles of any duration, is called "transpiration" cooling. The liner is surrounded with cold fuel, as with active cooling, but some small portion is diverted through pores in the liner, and evaporates on the hot side, protecting the liner from the hot reactive gasses. The problem here is that in order to work properly, transpiration cooling essentially requires a microscopically fine array of thermostatically controlled fast reaction valves, (Combustion in a rocket engine is turbulent, heat loading isn't just high, but chaotic as well.) to deliver to each point only as much fuel as is needed to accomplish the cooling. A very difficult feat to achieve with conventional engineering, but should be a cinch with nanotechnology. The nice thing is that you could also incorporate sufficient instrumentation to be confident the engine was in good condition, and potentially could even make it self-repairing to some extent.
So, that takes care of cooling, and diamond DOES have a very high thermal conductivity. Transpiration cooling is not as reliant on thin liners as active cooling, so tensile strength isn't a real constraint. You couldn't make a rocket engine out of utility fog, (Well, maybe an ablative engine.) but diamondoid blocks should be no problem. You would probably want to passivate their surfaces with something like saphire, though.
Joint strength >=50%, not 10% by Chris Phoenix, CRN 13:37, 3 Oct 2004 (CDT)
Tom, high joint surface area is the reason I invented the ridge joints. Read Primitive Nanofactory section 3.2.1, the paragraph starting "The width of the shim..."
Dupe time doesn't include pre-manufacture by Chris Phoenix, CRN 13:46, 3 Oct 2004 (CDT)
Tom, so far I have hardly thought about pre-manufacture. The reason is that I wanted the system to be as simple and programmable as it could possibly be. You're right that a production system would want to take advantage of that. But speed running flat-out is also important; what if you want to build something too big to pre-stock?
Molecular mills would likely improve efficiency quite a lot too, and I haven't included them either.
Chris
Rocket nozzles, propulsion, etc... by Chris Phoenix, CRN 14:45, 3 Oct 2004 (CDT)
Brett, at this stage of the game, wouldn't it be just as good to show that you could make other high-performance propulsion systems? For example, use a high-density fuel cell and/or solar cells to generate electricity to run an ion engine with a high-density ion injector.
Or, here's a crazy suggestion: mechanically throw the reaction mass out the back. Rocket exhaust, says Google, is generally limited to 4000 m/s. That's higher than the bursting speed of a diamond flywheel (or gear), but lower than the speed of sound in diamond (10,000 m/s). So make a rod that oscillates compressionally, and at maximum compression put a drop of reaction mass on it. I don't have time right now to figure out how to calculate how fast the rod will move and how much energy it will dissipate. I think you might have to slow it by adding a lump of mass to the free end, so the spring section will be just a fraction of the mass. And I strongly suspect that some dissipation mechanism will make this unworkable in practice.
Anyway, back to rocket nozzles: any eutactic material, by definition, doesn't wear. The only thing that would make it become non-eutactic is reactions between the exhaust gas and the surface (plus random radiation damage). As long as cracks can't propagate from the surface to the eutactic structure (perhaps due to pre-placed expansion cracks terminating in prestressed regions) then it should be possible to make a nozzle wall quite a bit thinner than today's versions, thus more flexible and with better heat conductivity. And diamond is an excellent conductor of heat anyway. So I think an all-diamond rocket engine with active cooling should work just fine.
Mass driver by Selenite 19:39, 3 Oct 2004 (CDT)
Chris, your propulsion concept falls in the "mass driver" category. The ones I've seen use an electromagnetic track to accelerate the reaction mass, but doing it mechanically should be perfectly feasible.
An array of very small mass drivers spread over an asteroid would be an interesting concept. You'd have redundancy and the ability to use differential throttling for attitude maneuvers. If your asteroid-processor is self-replicating you could have a steadily increasing thrust level as your produce more units. That'd give you a mining or Earth-defense application. Might make for a good cover illustration on the final report.
Propellants by Selenite 11:42, 6 Oct 2004 (CDT)
For a diamondoid rocket engine your best propellant combination is probably kerosene / hydrogen peroxoide. The peroxide is less likely to burn the tank and can be stored at room temperature. You do have to be careful with the surface chemistry to make sure there aren't any catalysts, and peroxide will decompose over long term storage, but it still gives decent performance without being toxic or cryogenic.
(no topic) by Tom Craver 20:54, 6 Oct 2004 (CDT)
Chris - the mechanical rocket scheme is interesting. Instead of "throwing" mass, what if you squeeze it out from between two laterally collapsing planes of diamond? Simpler mechanically, I think, than a longitudinally vibrating rod, and possibly with higher performance.
E.g. if there were a V between two adjacent planes of diamond with a 10:1 ratio of width of the opening of the V to the depth of the V, and if the V is completely closed at the speed of sound in diamond, it seems like the reaction mass (assuming it all started at the "bottom" of the V) should be squeezed out at 10x the speed of sound in diamond? Exhaust velocity of 100000m/s would be, to say the least, pretty excellent. I'm sure there are factors that would cut into this.
Powering it would be a matter of setting up laterally resonating standing sound waves, so that any energy not imparted to reaction mass is efficiently reflected for the next cycle - just keep pumping in energy at the rate that the escaping reaction mass damps the vibration.
There might be a problem with undesired bonding of reaction mass to the "nozzle" however? Perhaps if the nozzle were lined with a layer of the same atom as it uses for reaction mass - hydrogen seems ideal, as it couldn't build up in layers to clog the "nozzle" as other atoms with more bonding opportunities might - at least if my rusty memory of chemistry serves.
(no topic) by Tom Craver 21:25, 6 Oct 2004 (CDT)
Chris: Have you thought beyond the cube shape? E.g. trapezoids might be interesting - alternate them for flat, tightly joined surfaces, or put one or a row out of sequence to introduce some curvature - at the cost of needing some ability to tolerate gapping. Or you could use cubes for space filling, and "plate" some surfaces with a small variety of wedge shapes to provide for less friction. (Normal surfaces are pretty random and can probably bump along ok with 200nm features, but cube-based surfaces might tend to lock-up simultaneously across an entire surface if oriented just the wrong way - creating high peak forces, perhaps setting up ultra-high frequency vibrations in continuously sliding curved parts like bearings, or maybe tearing certain blocks loose?)
Kerosene/peroxide = less performance, but OK by Chris Phoenix, CRN 21:31, 6 Oct 2004 (CDT)
I thought of peroxide, but was hoping to get by with LOX. But after all, peroxide will get Black Horse to orbit. Given the probable ease of designing morphing air/spacecraft (bigger wings for efficient flight to altitude, folding up for higher speeds), and efficient rockets (morphing engine bells), and hybrid air/rocket designs, (retractable engines/ducts) it probably wouldn't be too much of a pain to use peroxide.
Electric supersonic "jet" compressor engines? by Chris Phoenix, CRN 21:53, 6 Oct 2004 (CDT)
I've been reading about turbojet engines for supersonic flight speeds, and it looks to me like you don't actually need to burn fuel in them if you have very compact electric motors and fuel cells. Can someone check me on this? The point of a turbojet engine is to:
- Slow the air efficiently to subsonic;
- Compress the air;
- Add fuel to heat the air, adding thrust and driving the compressor;
- Expand the air efficiently out the back, so it goes supersonic like in a rocket nozzle throat.
But what if you skipped step 3? You could still slow, compress, and expand the air, adding energy to it that would come out as speed. At higher speed its pressure would be lower, so the exhaust would be smaller than the inlet. But the inlet would presumably see a low pressure due to the compressor sucking air away from the shock wave, making its angle shallower.
Why would one want to do this? Efficiency. A jet engine is a heat engine, limited by Carnot. But fuel cells are not...
Lest we be confused by today's electric motor and fuel cell power densities, note that electrostatic motor power density scales inversely with size, so nanoscale motors might reach a petawatt per cubic meter. And fuel cells should be quite mass-efficient with better-constructed membranes.
Beyond cubes... by Chris Phoenix, CRN 21:57, 6 Oct 2004 (CDT)
I have been assuming that truncated cubes would be workable, as long as they had enough of the appropriate surface for handling and joining to other cubes. I don't think this adds much to the assembly planning.
So I've thought about thinking about other shapes, but I don't think there's much need for them.
Supersonic jets by Selenite 23:16, 6 Oct 2004 (CDT)
Hmmm. If you're not going to have combustion you might be better off skipping the compressor and just having a set of turbines to accelerate the already-supersonic air. There's some boats that use that principle example.
(no topic) by Tom Craver 12:33, 7 Oct 2004 (CDT)
Rather than go for fastest possible exhaust velocity (at least while in atmosphere), it seems like you really want to maximize the volume of air accelerated, rather than go for the fastest exhaust velocity possible.
For example, for a given amount of energy: E = (0.5)(m)((4v)^2 - (v)^2) == 0.5(5m)((2v)^2 - (v)^2) : same net energy input, the first increases air velocity by 4x, the second by 2x, but the second case moves 5x the mass. But in momentum: m(4v-v) = 3mv < 5m(2v-v) = 5mv : 1.67x more momentum for the same energy with the second case.
Meantime, every unit of frontal area you convert to "engine intake" reduces drag, which is also proportional to the square of velocity. So the ideal jet (or car) would have most of it's frontal surface covered with air compressor intakes. This likely isn't very practical with current technologies, but might be with nanotech.
Chemical-powered mass driver by Selenite 10:10, 12 Oct 2004 (CDT)
does this mean that a chemical-powered mass driver can be more efficient than a rocket engine?
Good question . . . depends on how you define "efficient" I suspect. You can get a better specific impulse since mass drivers don't have the inherent limit of chemical rockets. The system performance will depend more on how you deal with your fuel cell reactant mass. Storing or venting it would destroy your mass fraction, since you'd need to pack along additional reaction mass. The best bet might be to make the reaction mass of buckets to hold the used reactants, along with a couple of "starter" units. But this means each mass driver shot is only putting as much energy in to the mass as it would get if the reactant was burned. So your system efficiency would depend on what speed the mass driver can get the bucket up to, and how the dry mass of the fuel cell/mass driver compares to the equivalent rocket engine.
Various propulsion topics by Chris Phoenix, CRN 11:15, 12 Oct 2004 (CDT)
1) What's the difference between a compressor and a turbine? I've read that a flat plate is the most efficient airfoil at supersonic speeds; does that mean that simply spinning angled flat plates (vanes) in a supersonic airflow will accelerate it efficiently?
2) In the chemical-powered mass driver, I was assuming the spent propellant would be thrown overboard (in buckets, if you like) as the reaction mass. But I'm realizing this may not be an efficiency win after all. A rocket engine is a heat engine; but if the cold side is at 4 K, it'll be pretty darn efficient, hard to beat even with fuel cells.
The mass driver also has the advantages of low temperature (but the disadvantage of high speed) and scalability.
Spring-loaded supersonic squirt-gun rocket! by Tom Craver 14:59, 3 Dec 2004 (CST)
Spring-loaded supersonic squirt-gun rocket by Tom Craver 15:33, 3 Dec 2004 (CST)
Trying again... More on the idea of squirting liquid from between collapsing planes of diamond...
If the sound moving through the diamond is approximately a sinewave, the gap might not close as fast as desired. In that case, the wave could be re-shaped by adding "chocks" along the sides to keep the gap open until the compression wave has peaked, maximizing the closure rate when the chock is removed.
The gap would close at the speed of sound away from the removed chock, so that the sides begin to collapse together at the same time, with the collapse propagating inward over the surface of the gap at the same speed, creating a closing cone around the liquid.
No beating a rocket unless the energy source is non-chemical. by Brett Bellmore 12:26, 5 Dec 2004 (CST)
You're trying to supply your vehicle with momentum. Because momentum is mass times velocity, and energy is one half mass times velocity squared, if you're using chemicals to generate the energy, you're best off using them as reaction mass, too; Any part of the fuel that's not used as reaction mass reduces the momentum achievable per unit of energy.
It's hard to beat a rocket engine, if what you're doing is extracting chemical energy, and using it to accelerate those same chemicals. Rocket engines are pretty darned efficient, and have VERY high power to weight ratios.
Now, if you're in a low to moderate thrust regime, and have some non-chemical energy source available, (Nuclear, solar, beamed energy of some sort.) you have the potential to use non-thermal means to eject the reaction mass at higher speeds than can efficiently be done by thermal rockets. In a high thrust regime, you'd probably be better off using that beamed power to directly heat the reaction mass for thermal rocketry. (Which has the advantage that the reaction mass can have a low molecular weight, for better thrust at a given temperature.)
Beating rockets by Tom Craver 14:04, 8 Dec 2004 (CST)
The nano squirtgun is a rocket, of course, and yes, it definitely needs beamed or maybe nuclear power - with the objective of maximizing the payload mass fraction by maximizing momentum gained per unit of reaction mass.
I'm not sure whether thermal rocketry could match the exhaust velocity of the squirtgun - you'd need to be able to flash-heat the reaction mass extremely rapidly, or keep it bottled up under really extreme pressure and temperature conditions. Can one make a rocket that can endure significantly more extreme conditions than existing chemical rockets?
The squirtgun creates similarly extreme conditions - but only very briefly. One approach that might let it tolerate those conditions is to give the diamond planes a layer of hydrogen bonded to the surfaces, with that layer partially ablated during reaction mass ejection, but renewed in the next instant from hydrogen "caught" between the planes.
