Is it reasonable to build real world bigger ion engines?

[Nuke]

i was under the impression that natural uranium could be used in a reactor. however if thorium turns out to be easier to process in situ i have no problem with one being used over the other. it also provides a solution for the reactor launch problem. if we can minimize irradiation risks by doing our nuclear fuel mining processing off planet then its probibly a good idea to do so. its one of the reasons im more of a proponent of moon base rather than mars landing. mars might be a stunt but moon unlocks the universe.

[Kryten]

i was under the impression that natural uranium could be used in a reactor.

A civilian power reactor yes, but you won't get something with the power density required for space use. Producing natural uranium in the first place isn't too easy either.

[NERVAfan]

In order for Dawn to properly operate one of the engines we're talking about here - for example Ad Astra's VF-200 VASIMR (200kW, variable but typically 5N, 5000s) or ESA's DS4G ion thruster (250kW, 2.5N, 19300s) - while out there in the belt, it would need to carry solar panels rated in excess of a megawatt of power at 1 AU. You could potentially implement something like that from a pure technical feasibility standpoint, but I believe we all agree that it's absolute crazy-talk from a cost and complexity standpoint

It's certainly more expensive than Dawn's engine and solar panels, but I'm not sure it's that crazy if you use the "expanded IKAROS" model I'm talking about. The solar array is super-thin-film (IKAROS's was 25um thick) on an even thinner backing, and is deployed by spinning the spacecraft. I don't see why that's massively complex.

A megawatt at 1 AU is (assuming the solar panels are even 10% efficient) something like 7400 square meters of solar panel, or 86 meters on a side. I don't think that's all that crazy for these very thin films (although the wiring to carry the power will add mass).

Here's the JAXA brochure I was referring to (it's at the bottom of the second page):

http://global.jaxa.jp/activity/pr/brochure/files/sat28.pdf

"Next plans

The second mission will take place in the late 2010s. It will involve a large-sized solar power sail with a diameter of 50m, and will have integrated ion-propulsion engines. The destinations of the spacecraft will be Jupiter and the Trojan asteroids"

[NERVAfan]

i was under the impression that natural uranium could be used in a reactor. however if thorium turns out to be easier to process in situ i have no problem with one being used over the other. it also provides a solution for the reactor launch problem. if we can minimize irradiation risks by doing our nuclear fuel mining processing off planet then its probibly a good idea to do so. its one of the reasons im more of a proponent of moon base rather than mars landing. mars might be a stunt but moon unlocks the universe.

Asteroids have much lower delta-v to launch from than the Moon, though. If you're looking for an outpost to make other destinations easier, asteroids are better than the Moon.

[Streetwind]

The problem is that a space based nuclear generator gets maybe 10% efficiency at best, because it can't get rid of the heat easily, and the insane mass of the cooling and power conversion systems. If we take the SNAP8 as an example, it weighs 4.5t, produces 600kW thermal but only 35kW electric. For comparison, NERVA had an empty mass of 35t for 1100MW, 2000 times more powerful but only 7 times heavier.

I couldn't find how much the actual SNAP8 reactor core weighs, but it has 6.5kg of uranium, and 12 control rods weighing 12kg each, in a steel core that's 23cm in diameter by 53cm height. The coolant alone is more than 400kg, and the heat exchangers and piping probably account for most of the mass, so I'd be surprised if the reactor itself was heavier than 200kg. On the other hand, you have to take the turbopump and and nozzle.

Let's do some math, boiling and heating 1kg of H2 to 2500K takes 18MJ, with 600kWt available, it means a flow of 32.5g/s. I doubt the pump and nozzle required to handle 32g/s of hydrogen have masses in a the hundreds of kg.

If we assume a 600kW NTR would weight 1t, 8.3km/s ISP, a 1t payload and a 6km/s deltaV budget, we would need a wet mass of 4.1t.

Using the SNAP8 and a ion thruster of negligible mass and 100km/s ISP, we have a dry mass of 5.5t, and a wet mass of 5.85t.

(Psst... don't say "Isp" when you mean "effective exhaust velocity", it confused the heck out of me for a moment )

The SNAP8 was deployed for testing in 1963 - that's more than half a century ago. It's somewhat of an unfortunate example. With 35 kW usable electric power out of its 600 kW heat production, that means it must reject 565 kW worth of heat from radiators. Its conversion efficiency is indeed around 10%. Now compare the SAFE-400 I talked about in post #17 in this thread. That one was designed and built in the 2000's and pulls 100 kW electric power out of 400 kW reactor power, leaving only 300 kW heat to be removed. That's almost 2.5 times the efficiency of the old SNAP8, almost 3 times the absolute power output, and requires only roughly half the radiator power... which, with advanced in technology in the past 50 years, probably would weigh less per W of heat rejection as well. The entire SAFE-400 system weighs just 0.5 tons (without radiators, admittedly, but including the closed brayton cycle heat engine) and has the same dimensions as the SNAP8 reactor core.

Fission reactors really have come a long way, we just don't generally hear about it because the general public is allergic to the topic. And the SAFE program isn't even official - NASA scientists built that thing in their spare time, outside of work hours, with "discretionary funds" (in other words, they plunder the tip jar and compete for unassigned leftover lab funding). Imagine what an official, fully funded project could do.

There is, of course, the question how a hypothetical modern-day NTR would perform in comparison, given the advances in reactor technology. I unfortunately can't answer that.

It's certainly more expensive than Dawn's engine and solar panels, but I'm not sure it's that crazy if you use the "expanded IKAROS" model I'm talking about. The solar array is super-thin-film (IKAROS's was 25um thick) on an even thinner backing, and is deployed by spinning the spacecraft. I don't see why that's massively complex.

A megawatt at 1 AU is (assuming the solar panels are even 10% efficient) something like 7400 square meters of solar panel, or 86 meters on a side. I don't think that's all that crazy for these very thin films (although the wiring to carry the power will add mass).

Here's the JAXA brochure I was referring to (it's at the bottom of the second page):

http://global.jaxa.jp/activity/pr/brochure/files/sat28.pdf

"Next plans

The second mission will take place in the late 2010s. It will involve a large-sized solar power sail with a diameter of 50m, and will have integrated ion-propulsion engines. The destinations of the spacecraft will be Jupiter and the Trojan asteroids"

It's an interesting enough proposal to be sure, but do you know the W/kg rating of that solar sail when used for photovoltaics? I couldn't find anything to that effect in the brochure. That's the key performance indicator you need to look at when deciding whether or not it's competitive with a nuclear electric power source, especially that far away from the sun. It would literally be the first mission ever that relies on solar power beyond the belt. Everything else we've shot out there (Voyager, Cassini, New Horizons...) relied on RTGs as far as I remember.

Edited October 2, 2014 by Streetwind

[Nibb31]

Asteroids have much lower delta-v to launch from than the Moon, though. If you're looking for an outpost to make other destinations easier, asteroids are better than the Moon.

It depends on the orbit of the asteroid. Asteroids don't stay in one place.

You would need a lot of delta-v to rendez-vous with the asteroid, and then a lot of delta-v to go from the asteroid's orbit to your destination's orbit. I don't see how this serves any purpose.

[NERVAfan]

It's an interesting enough proposal to be sure, but do you know the W/kg rating of that solar sail when used for photovoltaics?

No, but I think that one may be mainly a solar sail with just some of the area covered.

We know that IKAROS' thin films were 25um thick

It would literally be the first mission ever that relies on solar power beyond the belt. Everything else we've shot out there (Voyager, Cassini, New Horizons...) relied on RTGs as far as I remember.

First to use solar power for electric propulsion, yes, but not first to use solar power at all; Juno is a solar-powered mission to Jupiter.

[NERVAfan]

We know that IKAROS' thin films were 25um thick

Argh.

What I meant, but forgot to add, was:

An order of magnitude estimate is possible from that; assuming the cells are about silicon density of 2000 kg/m^3 and assuming 10% efficiency, you'd get about

20 m^2 / kg

~135 Watts/m^2 or ~2.7 kW/kg at 1AU

~5 Watts/m^2 or 100 W/kg at 5.2 AU (Jupiter's distance from the Sun)

I don't know how much mass the cables/wires to transmit the power would be, though.

[Streetwind]

First to use solar power for electric propulsion, yes, but not first to use solar power at all; Juno is a solar-powered mission to Jupiter.

*peers at Juno*

...So it is. I stand corrected. Quite impressive to pull that off with just 4% of the sunlight intensity available at Earth.

Still, with a peak output of as little as 420 W over the course of the mission, a good portion of which will go into self-heating to avoid failure, Juno is far away from being able to power even an NSTAR (much less a more modern electric thruster). It would have to have at least six times the panel area, increasing the mass of the entire spacecraft by about 50%. And that's assuming it can use 100% of the panel area during burns, which is unfortunately almost never the case. At that point, even keeping dV the same by reducing fuel carried, the electric propulsion variant of the spacecraft would be noticably heavier in total mass than the traditional propulsion variant they currently use, while at the same time dropping main engine thrust from 645 N down to 0.09 N.

I really do think that a simple hypergolic rocket engine is by far the better choice for that mission and remain unconvinced of the utility of solar electric propulsion that far away from the sun

[NERVAfan]

It would have to have at least six times the panel area, increasing the mass of the entire spacecraft by about 50%. And that's assuming it can use 100% of the panel area during burns, which is unfortunately almost never the case.

That's why you need to use thin films, not traditional solar panels, for electric propulsion beyond the asteroid belt.

I really do think that a simple hypergolic rocket engine is by far the better choice for that mission

Well, certainly, because the total delta-v needed is small enough that chemical propulsion is feasible. This would only really be useful if you were doing a very high-delta-v mission that would require unworkable mass ratios with chemical rockets.

and remain unconvinced of the utility of solar electric propulsion that far away from the sun

I think it might end up competitive with nuclear at least out to Jupiter given the existence of thin films, and maybe workable even at Saturn someday if nuclear remains politically difficult.