I think the trick is to refrain from trying to use the ion thruster the same way we use a chemical one. Instead of diving into Neptune's gravity well with a hyperbolic excess of over 4 kps, we want to gently kiss the Hill Sphere, and almost immediately go into a huge, slow circular orbit, from which we gently spiral down.

Trouble is, the math for that is a bit challenging -- sort of just doing a numerical simulation. Not sure anyone has actually worked this out.

Anyway, I think that disposes of most of the diference between Neptune and an asteroid. If the numbers work out, anyway.

--Greg

I've seen transfer orbits to Mars with low and constant thrust, I think even from the '60s. It boiled down to first accelerating, then decelerate, and then accelerate again, and you'll end up with about zero relative velocity. Not sure how long it will take though, but maybe on long distances like Neptune the low thrust option may be quicker.

Bruce Moomaw is currently writing a very interesting series of articles on the exploration of the non-Martian Solar System (his words). His latest article has some news about the Neptune Orbiter as the proposed 4th NASA flagship mission.

http://www.spaceblogger.com/reports/NASA_C...eptune_999.html

>Trouble is, the math for that is a bit challenging

One approach that might simplify this is to look at it in terms of changes in energy rather than orbital mechanics, at least for relatively circular orbits (or locally circular spirals). Assume that the energy expended by ion drive exhaust translates to changes in potential energy (equivalent to orbital radius) with regard to the body being orbited, assuming that the thrust of the ion engine is relatively orthogonal to the gravity vector. I recall seeing some elegant solutions to a variety of problems dealt with in this manner.

Bruce's article describes a Neptune orbiter mission with a 15 year transit time to Neptune based on a similar idea to the one that was developing here, although his example does deal with more realistic mass scales and tech that is very close to being available. If you take a relatively low mass craft that you launch outwards with a significant initial chemical kicker, then use a Radioisotope Electric Propulsion system (ie something like an RTG+Ion drive) to slow you down in the latter half of the mission and then use a small chemical engine for insertion braking that would require only about 10% of the final orbit mass for braking fuel. The model he describes is a 500kg dry weight orbiter, ~400kg of Ion fuel most of which will be used up in the approach deceleration phase and a 750watt Stirling RTG power source (weight about 250kg), leaving you ~250kg to build the chasis, comms, command systems and the instrumentation.

_Assuming_ the components could be scaled down, the orbital equations should scale linearly with mass and this would indicate that ~1kg system could deliver a 100g "probe" to that range in ~15 years without breaking any fundamental laws. I absolutely accept that it is fantasy for now but I can only hope that something like this will eventually be built.

The difficulty with that idea is that it is just as hard to get down into a gravity well from a high circular orbit as to is get out of it up to the orbit. On the other hand if you get into a highly elliptic orbit you get to see both the near and far parts of the circum-neptunian area and you can still use the ion thruster to adjust the perineptunium and the orbit generally.

True, and it actually requires a good bit more delta-v to do it via a spiral, BUT, it doesn't take 10x as much, and the ion drive does have 10x the specific impulse of a chemical thruster.

Using the ion drive to get into a highly elliptical orbit on purpose is a bit more challenging. (Mathematically, I mean.) Better would be if you could use aerobraking to kill 1/2 kps or so at Neptune and then let the Ion drive raise your periapsis enough before the next pass, but figuring out a reasonable amount of aerobraking to expect is another hard problem.

On the Van Allen Belt front, I note that using a chemical rocket (309 s specific impulse) to go from LEO to 5,000 km up, you'll lose roughly half your payload. So if you started with 28 tons in LEO, you'll only have 14 at 5000. If you used an ion drive, on the other hand, you'll keep essentially all of it. So my question is, would you really need 14 tons of shielding if you opted for a lesiurely trip through the belts?

--Greg

I don't know if this was mentioned earlier, but there is a good description for a Neptune orbiter mission in the Decadal Survey white paper on the Exploration of the Neptune System:

http://www.aspbooks.org/publications/272/neptune_final.pdf

To return to a question nprev asked earlier about data rates at Neptune.

This sort of stuff is a long, long way from developing into a space rated system that could be used for S/C comms but it does describe 24Ghz and 77Ghz solid state phased array antenna systems with total gains between 41 and 61dB from solid state SiGe arrays with total physical dimensions < 1 cm^2.

For the sake of argument lets take MRO's X-Band comms as a baseline and work down and out.
At max Earth/Mars range (400m km) MRO gets about 650kbps using a 100 watt amp and a 3m parabolic antenna (~ 47dB gain).

(Assuming that overall gain and efffective data rate are linearly related which is more or less the case)

Drop the Tx power to 1 watt, 100x loss => 6.5kbps.
Change the antenna from 3m Parabolic (47dB) to a solid state phased array (41 dB gain), ~4x Loss => 1600bps.
Change due to the range at Neptune (4600m km) ~132x Loss => 12bps.
Change to Ka Band. ~4x gain => 48bps.

So my 100bps WAG wasn't far off. A spacecraft antenna gain improvement to ~60dB would raise that to 5kbps and if the DSN was updated to arrays of 400 x 12M antennas we would see a further 10x gain which would bring things to 50kbps. All of that from a base transmitter at Neptune that is about as powerful as a cellphone.

Comparing this approach to New Horizons comms seems like a smart idea since we know what those should be. NH will give us 300-600bps at Pluto using a 12watt Tx amplifier and a 2.1m antenna.
So start with MRO's 650kps, 100watt, 3m parabolic at 400m km.
Drop the Tx power to 12 watt, 8x loss => 78kbps.
Change the antenna from 3m to 2.1m Parabolic, 3dB/ ~2x Loss => 39kbps.
Change due to the range at Pluto (~4600m km) ~132x Loss => 295bps.
Looks right to me.

Back to the propulsion question, one of the CubeSats was to have tested a Micro Vacuum Arc Thruster, which might be relevant, though I think not as efficient as the larger Xe ion drives as used in Smart 1, etc., see

http://cubesat.atl.calpoly.edu/media/Docum...uster_paper.pdf

If you can pull-up reference 3 of that paper (in Review of Scientific Instruments, if you have access), it states that "scaling existing electric propulsion engines such as Xe ion engines ... down to 1-10 W power levels is not practical. The unavoidable overhead of propellant storage tank, flow controls, and plumbing in Xe ion engines ... makes their overall efficiency unacceptably low at these power levels." (but they don't cite a reference).

These scaling issues strike me as directly addressable via MEMS/Microfluidics. I'm not sure about a fuel tank for Xenon - I gather this is normally stored as a liquid under high pressure. Am I incorrect in recalling that it becomes easier to store something at high pressure as the volume becomes smaller? I would envision something like a 10 cm carbon fiber composite tank as a feasible option, or perhaps several smaller tanks for redundancy.

It sounds like helvick has addressed the power & communications issues. Seems to me like this notion is sounding progressively more feasible, particularly if the probe can piggyback to geosynchronous orbit with a comsat launch. Construction + launch ought to be in the $100K to $200K range, which sounds amenable to donations.

I was surprised by Bruce's implication that New Horizons was comparable to Cassini in terms of its scientific capabilities. I seem to remember that Cassini weighed about 6 tons at SOI, half of which was fuel (and half the fuel was spent on SOI), while NH weighs a bit under half a ton. I know technology has gotten better, so I'd expect us to be able to make equally capable probes for less, but is it THAT much less?

Do any of our experts have a ballpark guess as to how light a Cassini-equivalent probe would be if we built it with modern technology?

--Greg

At the risk of going waaay off topic here I just have to wax lyrical on just how dramatic the changes have been over the 10 years that elapsed between the launches of Cassini and New Horizons. To put the capability delta in perspective it is worth looking at just how much other things have changed -

General purpose consumer CPU's
1997 - Averge CPU speed 200Mhz single core. Pentium II andPentium Pro.
2006 - Average CPU speed 3Ghz, dual core. Dual Core P4 CPU's
Delta - 30x.

Consumer PC Hard Drive capacity.
1997 16.8GB (initial introduction of 16.8GB IBM Desktstar using Giant Magenetoresistive hHeads)
2006 500Gb
Delta 30x.

Solid State Memory - Compact Flash
1997 - 64MB
2006 - 4GB
Delta 62x

Digital Cameras
The very first megapixel resolution consumer cameras arrived at the very high end of the market in 1997 ($1000+). Just think for a second how much that technology has changed and then think how fantastic the Cassini pictures still are given the fact that the cameras predate those 1997 megapixel clunkers by a few years.

Cellular Telephony
The progress in the cellular telephony market should also give you some idea of the scale of the change in the sophistication of technologies in the last decade. The improvement in the basic capabilities of cellular handsets in the intervening period is hard to describe - in 1997 most countries had a cellular market penetration well below 20% however by 2006 it was at 90% or higher for more than half of the planet including much of the so called 3rd world with annual handset sales exceeding 1 billion units for the first time. In 1997 SMS messaging was a novel technology that few people had ever even heard of but by 2006 it was a technology in daily if not hourly use by the majority of the population in Europe and Asia with the Americas catching up fast. In 1997 a handset's "standby" battery life was typically 2-3 days but by 2006 10+ days standby time was not unusual for similar devices. In 1997 reliability was suspect at best and expected connectivity SLA's were in the region of 95% for first time connection success, by 2006 that had exceeds 99.9+% for most mature markets. Between 1996 and 2007 costs per unit call have dropped by a factor of 10 or more while the scope of services has increased dramatically (now voice + high speed data costs << voice alone cost in 1997).

Digital Music
In 1996 very few people realised that you could actually record\copy\process\listen to music using a computer, personal digital music players _did not exist_. The MP3 format was first patented in the US in 1996. By 2006 there were more than a billion personal digital music players in daily use.


Conculsion
Personally I'm surprised that New Horizons is so big.

( Only joking. John\Alan and the team, I think she's the bomb really).

This doesn't make sense to me. It's decades since I did any solid materials\fluid dynamics work but I'm certain that a linearly scaled smaller pressure vessel will be able to handle progressively higher pressures.

That said it is true that it is harder to build a 1cm diameter fluid management system of a given complexity than (say) a 10cm diameter version of the same thing but the reason for that is that mechanical precision gets harder as machining sclaes diminish not that pressure systems become problematic as scales diminish.

MEMS\Microfluidics should be able to provide a solution to all of these problems someday. The general impression that I get is that folks think this is all pie in the sky (at leat for now) but it is worth noting that the microfluidics capabiliies that have been perfected over the past 15 years of Inkjet system development are more than capable of managing a fuel delivery system with the precision and finesse that we require for this sort of engine. Single nozzle control in the pico-litre ( 1E-12 ) scale is a reality today for example. The problem is that all our current technology is geared at doing this for un pressurized room temperature fluids at the earths surface and not for pressurized low temperature fluids in space\vacuum.

Boy, does THAT sound like a fruitful grant proposal topic for one or more people in our forum (hint, hint.... ) This might be a key enabling technology for UMSF that would have extremely broad applicability...too bad I ain't smart enough to invent it & get embarrassingly rich off of the patent...

Algorimancer and nprev: What am I missing here? Why is it desirable to scale the ion engine down below 10W? It seems that the engines only weigh about 8 kg now. Or is the issue with what the power source weighs?

On that topic, I do think there's way too much angst over using nuclear power for this, largely sparked by the kooky anti-Cassini protests of a few years ago. I note that New Horizons didn't draw any protesters to speak of, and I doubt any subsequent missons will either.

--Greg