Re:So why is Gentoo the right choice for this?
on
Embedded Gentoo?
·
· Score: 4, Insightful
How many program projects that you know of offer linux-ppc or linux-mips or linux-arm binaries?
Well, Debian for one. In fact, Debian supports x86, Motorola 68k, SPARC, Alpha, PowerPC, ARM, MIPS, PA-RISC, IA-64, and S/390 architectures. Porting to the AMD64 and Hitachi SuperH is also underway. Note that the 68k, PowerPC, ARM, and SuperH are all popular for embedded applications.
Ok, seriously, are you just trolling here? Or are you really trying to equate a loss of a life-saving antibiotic with the "life and death" prospect of missing the next episode of 'Survivor'?
My point wasn't that what the networks are doing isn't bad, but rather that your reaction (the raping, killing, and burning) was a little out of proportion
Humanity survived for thousands of years without TV, movies, or DVDs. I'm sure we can do it again if things get as bad as you suggest.
If it gets as bad as you are suggesting hopefully people will stop buying media products (this is survivable - see above) until the products improve: the only reason that networks and studios do this stuff is because people continue to buy their products. Stop buying, and they'll change their product until you do buy.
Well, there is a CSP/Occam-derived add-on for Java that supports a robust concurrency model. It just isn't widely used (yet). Note that there are similar concurrency toolkits for C and C++ (in fact I believe that the x86 port of Occam is built on top of the CCSP toolkit).
The sad thing is that the Transputer is still a great idea - it died mostly due to poor business decisions by INMOS (late getting newer processors out). I mean, look at the popularity of SMP, Beowulf clusters (that perenial/. favorite), and the like. The transputer was specifically designed to support that kind of thing in a clean way, and had a language (occam) that made it easy to program for multiprocessor systems. Sigh. Sometimes it seems like the computer industry has a habit of ignoring great innovations, and then reinventing them badly a few years later.
It's probably worth adding that from the TFA it seems that VxWorks code is shared between different spacecraft programs. So the folks who can test on real satellites are sharing their patches, fixes, and features.
There are several open source RTOS's out there, if you want to provide some kind of educational aid. Off the top of my head, I can think of RTEMS, eCos, and RT-Linux. I've seen several real-time courses use the MicroC OS, but I don't recall if it is open source or not. The odds of WindRiver (NASA doesn't own the code) open sourcing VxWorks are pretty minimal I would imagine.
Shielding does not protect against single-event upsets (particle-induced bit flips), it only provides some mitigation against total ionizing dose (which causes long term cumulative degradation as a result of drift in transistor operating parameters). There are design techniques and fabrication processes that can reduce the likelihood that a circuit will suffer upsets, but it's still standard practice to provide either redundant memory, or error detection and correction coding. In the case of MER they had 3 physically separate PROMs carrying identical copies of the flight software, and the RAM was (IIRC) protected by an EDAC code implemented in a rad-hard FPGA.
You are confusing Mars Pathfinder (a 1997 mission, which suffered a priority inversion problem) with Mars Exploration Rover (a 2003-2004 mission which suffered a file allocation issue in Flash memory, and the subject of TFA). Although both used VxWorks.
NASA has had an active formal methods/formal verification program for a number of years, located at NASA Langley. They mostly do research, but have worked on a few practical applications, mostly in the shuttle program. Additionally, JPL recently (2003) set up the JPL Laboratory for Reliable Software, which is chartered to look into formal verification among other things. The lead technologist in the LaRS is none other than Gerard Holzmann, the man behind SPIN.
Having said all of that, I'll agree that formal verification at NASA is in its infancy, and is facing an uphill battle for acceptance (witness how long the Langley group has been trying to push formal methods). It'll be interesting to see what happens with JPL's LaRS.
Actually, there are some spectacularly rich people in rural areas. They just aren't all that showy about it. Similarly, many of the folks in LA that drive SUVs are living beyond their means, but obsessed with image.
Considering that the Chinese government builds and operates its own launch vehicles I doubt that launch costs are an issue. That said, your point about spatial resolution is bang on, and is the number one reason I can't imagine that they'd mess with a GEO "spy sat".
Spy sat fly low because the required camera resolution and spacecraft pointing accuracy is far lower. They fly in highly inclined orbits because that allows them to cover the most latitude - polar orbits will cover the entire world. They also fly highly-inclined because it allows them to achieve a sun-synchronous orbit that provides good lighting conditions for taking imagery.
And actually, the original poster was correct in his usage - geosynchronous simply means a 24 hour circular orbit (35,786 km altitude). Geostationary refers to a geosynchronous orbit with a (nominally - there is some drift) 0 degree inclination, which minimizes the amount of apparent north-south motion and keeps the spacecraft at what appears to be a fixed point in the sky.
They were also built with *1970s electronics*. Surely you, as a poster on slashdot, know how much electronics have improved since the 1970s.
The problem is that spacecraft power consumers consist of more than just electronics. And yes, electronics have improved since the 70's, but they also do a lot more than their 70's ancestors do. You will consume 60-70 W in comm system alone (transponders plus amplifiers), even neglecting high-rate transmissions. Your onboard computer (plus associated interface circuits) will consume upwards of 30 W (for a non-redundant system - more for a redundant system). Survival heaters will probably consume anywhere from 70 W up to 200 W (yes, you can use some isotope-based heaters, but they cannot be switched on and off, so they are of limited use). Between reaction wheels, star trackers, IMUs, and the like, your attitude control system will consume 70+ W at a minimum. None of this includes the inevitable inefficiencies in DC-DC conversion and power distribution. These are real estimates, based on real deep-space (as opposed to near-earth) missions (NEAR, DAWN, ACE, etc).
...less stabilization force needed, although the fact that they're in a non-uniform gravity well with a variety of other forces will somewhat negate this
"Non-uniform" gravitation has little or nothing to do with it. The prinicipal disturbance torques acting on a spacecraft are gravity-gradient (which has little to do with the non-uniformity of the field), magnetic, solar radiation, and aerodynamic torques. Aerodynamic torques are negligible even for high LEO orbits. GG torques will not affect a deep sapce mission. The other two are still a concern. Plus, you are constantly turning your spacecraft to keep the thrust vector pointed in the appropriate direction.
Not really. Why would you design a craft to constantly be expecting to receive when you have a delay time of hours? Scheduled command receipt times make far more sense, unless you've just got power to burn.
[sigh] Look, no one (or at least no one I am aware of) designs spacecraft that switch their command receiver on and off - there's too much chance that it won't get switched on again at the right time (or ever). Plus, if you are constantly thrusting you are going to want to get regular telemetry, just to amke sure everything's still operating correctly.
The fact remains that with 1970s tech, the Voyager craft idled on minimal power. Now you're trying to say that it's not realistic with 2000s tech. That rings incredibly hollow.
See above - we're talking at least 200 W, probably more. And these are not subsystem you can switch off: if you are constantly thrusting you need to be controlling your attitude - which means ACS on, flight computer on, comm system on for telemetry and command updates. You can't "idle" at 0 W, or even close to it.
If it's not enough for you, then we can upgrade the power supply; Cassini, after all, launched with ~750W of RTG power (of which a peak load is only needed during flybys) - and there, we're not even talking about a craft that powered its propulsion via electricity. If you want to devote even a small fraction of that saved propellant mass to RTG power, you can get a T-140 without having to launch an unreasonable amount of plutonium (i.e., the costs of the plutonium will still be insignificant compared to the costs of the entire project). Then, with your remaining propellant savings, you can fill it with insturments, and you can have your insturments make use of your higher power supply as well. It's a win-win situation.
The T-140 requires a minimum of 1.8 kW. If you're going to try to supply that much power (plus whatever you need for spacecraft power), why not just use a reactor? You are now attempting to fly almost of 500 kg of RTGs, versus a reactor that could weigh less than 500 kg, and produce significantly more power output (while using cheaper uranium, instead of plutonium). That's been my point al
Voyager 1 and 2 each need only 215 watts, and that was launched in 1977. You're saying that we've 2.5xed our idle power load since the 70s (I can't find the idle power req's for DS1; if you can cite a reference, please do)?
Voyagers 1 and 2 were deliberately built simple and rugged, and were tightly constrained by the power available from their RTGs. Modern spacecraft are more complex and often consume more energy. Spectrum Astro's DS1 factsheet gives the bus orbit average power as 500 W. Admittedly, that probably isn't with all systems "idled" (whatever that means), but there is a limit to how much you can switch off on a spacecraft in mid-flight.
I can run a graphing calculator for a day on a couple double a batteries that has 10 times the processing power of these craft
A graphing calculator is not an internally redundant rad-hard flight computer (30+ W), with its associated I/O interfaces. It does not have a comm system (which will draw power even when not transmitting - you always want to be able to receive). It does not have attitude sensors. It does not have attitude actuators (which can be large power consumers on a 3-axis stabilized spacecraft, as you would want to be with a low-thrust system since spin-axis precession is probably infeasible). These things all take power. 500 W? Not necessarily. Several hundred W. Quite probably.
That the cost of the plutonium is completely dwarfed by the cost of launching the fuel?
The cost is not just the cost of plutonium itself, but the additional cost and complexity that results from flying a radioactive (and toxic) substance. Cassini cost $1.2 billion just to develop - the launch cost is a small fraction of the development cost, let alone the total mission cost. Admittedly, not all of Cassini's cost was driven by the RTGs, but they definitely contribute to a higher cost than you would otherwise see.
Hall effect thrusters, while having ISPs of "only" around 2,500 (compared to ~450 for a good LOX/LH engine) give power/thrust ratios of near.65-.70 newton per kwH in the case of the T-220 (assumedly the T-40 performs similarly). A 200W T-40 HET, therefore, would well outperform DS1's engine operating at full power, and yet use a tiny fraction of the propellant of a chemical engine.
Unfortunately, electric propulsion system performance doesn't tend to scale linearly. There are economies of scale that come with larger systems. Quoting from the Pratt & Whitney website:
The T-40 operates at 0.1 to 0.4 kW and produces 5 to 20 mN of thrust, with specific impulse values varying between 1,000 and 1,600 seconds, depending on operational conditions.
a single Cassini RTG could power a T-40 HET with power to spare (you can bring more RTGs as you want for insturments) and produce over 100 millinewtons of continuous thrust. Over a year, it could impart, to a 1000 kg spacecraft, over 3000 m/s delta-V**. Seing as it took Gallileo 5 years to get to Jupiter, that's 15k m/s delta-V** - I.e., you could get to Saturn using *no* gravity assists as fast as Gallileo did *using* gravity assists.
Since you have overestimated HET performance you are getting numbers that are unrealistic (which is exacerbated by the exponential dependence of propellant mass on Isp). Using your numbers (which are optimistic) we get:
Isp 3800 s dv 15 km/s mdry 1000 kg thrust 0.1 N
mprop 496 kg TOF 5.8 years
However, if we use realistic (but still optimistic - I'm assuming the upper end of the quoted T-40 performance range, which requires closer to 400 W) numbers for low-power HET performance, we get**
Isp 1600 s dv 18.84 km/s mdry 1000 kg thrust 0.02 N
Where did you get your power ratios? Stirling cycle engines can get near the maximum Carnot cycle 50% efficiency - usually 30-45% efficiency.
From a NASA report describing their Stirling Cycle radioisotope generator - real efficiencies, not theoretical numbers.
Also, RTGs don't use Peltier-type devices (thermoelectrics) - they're actually thermionic. Thermoelectric ("Peltier-type") devices use a junction in electrically conductive/heat insulating materials. Thermionics use a vacuum between two extreme-heat-differential plates, where the hot-plate acts like a cathode and emits electrons.
I don't know where you are getting you're information from. There may be a few "RTGs" that use thermionics for energy conversion (I haven't heard of any though). The majority of RTGs do use thermoelectrics. For example, Cassini's RTGs make use of SiGe unicouples for energy conversion. Thermionics are much more common in space-going reactors, such as Topaz, Topaz II, and the US SP-100.
In-transit? You don't need much insturment power for the dozen or so years to the outer planets. What, do you think they're gathering nonstop radar images of deep space while flying?
It's not instrument power that's the issue. It's power for the comm system, the onboard computers, the attitude control system, and the thermal control system. Yes, you can reduce the amount of pwoer that these things draw, but you cannot make it zero, and we are still talking about a significant number of Watts.
As I mentioned, DS1 *did* use just a couple hundred watts for its ion engine for a good portion of its flight, so it clearly is a reasonable option. Hall effect thrusters use even less electricity for the same total amount of thrust, with a penalty of lower ISP (but still far greater than chemical rockets).
And as I mentioned, DS1 had a power source that produced far more than the amount of energy that it needed for the thruster - running both spacecraft systems and thruster off of "a few hundred watts" is not practical. The DS1 spacecraft, excluding the thruster, used about 400-500 W of power (see my points above about other subsystems that require power).
Why exactly would you not? Ion thrusters have plenty of time (months to years) to line up the craft for planetary assists. You might have trouble getting assists from moons when in orbit around a planet due to the short differences in time between encounters, but that's about it.
There is plenty of work being done on combing low-thrust trajectories and gravity-assist trajectories. I don't dispute that there is a lot of potential there. The question is whether or not the RTG-based design that you are proposing would be worthwhile in such a mission. I am saying no. See below.
Given standard launch cost rates (~25k$/kg for escape velocity), it would cost about 140 million dollars to launch Cassini. 78 million of that would go to launching just its propellant. RTG costs are trivial in comparison to propellant launch costs.
Actually, it cost closer to $450 million to launch Cassini, since they used an ultra-expensive Titan LV. The problem with your math is that launches aren't sold by the kg, they're sold by the launch vehicle. Would Cassini have been light enough to get to a smaller launch vehicle if you switched to ion thrusters? Hard to say - yes, you will get some propellant mass savings due to better Isp. But, either you use the existing power system, which means very low thrust, a seriously extended mission duration, and a greater need for redundancy, or you beef up the power system (use something other than RTGs). The latter approach being the one the JIMO has taken.
10x the ISP is *NOT* minor gains. 10x the ISP is a huge gain.
Yes, it is. But you are failing to account for the additional mass of the power processing infrastructure that is necessary to support an ion thruster, or the extra thermal radiators you will need due to the inefficiencies of
Hmm, actually, I was thinking that JIMO was scheduled to use an RTG, but they're looking to use a somewhat more complex reactor.
JIMO has, AFAIK, always been baselined to use a reactor. It's the only thing that makes sense. Particularly since they are talking in terms of hundreds of kW of power.
Apart from the fact that NASA is developing higher power stirling cycle RTGs
If it's Stirling Cycle it's not and RTG - RTG stands for Radioisotope Thermoelectric Generator, and uses Peltier-type devices to directly convert thermal energy into electrical energy. The drawback is that the conversion process is not very efficient (although it does avoid the need for moving parts). The reason that NASA is looking into Stirling Cycle Radioisotope Generators is that they can boost the energy conversion efficiency to maybe 20%. But you still only get a relativle small amount of pwoer out of a Radioisotope Generator of this kind (I believe that they are talking about 55 We from ~500Wth). AFAIK Europa Orbiter was a NASA (not ESA) mission that used chemical (not electric) propulsion, so it's not really relevant to this discussion.
Deep Space 1, for example, operated its ion drive for months at 520 watts
That's nice, but what are you supposed to run the rest of the spacecraft on? DS1 had solar arrays that produced ~2.5 kW at BOL, so even at max power on the ion thruster they still had a bunch of power left over to run the other spacecraft systems. I won't dispute that HET's are somewhat more power efficient, but not enough to make a significant difference compared to electrostatic ion thrusters.
I'm not arguing that low-power ion thrusters powered by an RTG won't produce more thrust in a Jovian orbit that a SEP system (although it depends a lot on the size of your solar arrays). I'm arguing that the thrust levels that you can achieve using any power source that produces only a few hundred Watts for both spacecraft and thruster is insufficient to perform a deep-space mission - the resulting trip times become unimaginably long. Now, perhaps if you were to combine traditional ballistic gravity-assists with RTG-based propulsion for say minor stationkeeping maneuvers you might be on to something. But somehow I doubt the (high!) cost and complexity of using an RTG would be worth the minor gains that you might make in propellant savings.
Uh, it's unlikely that anyone is going to use RTGs to power an ion engine. They simply produce too little power. For example, all three of Cassini's RTGs combined produce a mere 890 W (at BOL) - ion engines need on the order of several kilowatts to produce anything approaching a useful amount of thrust. Plus the plutonium needed to build RTGs is very expensive.
There is a lot of interest in so-called "Nuclear-Electric Propulsion" (NEP) right now, but no one is talking RTGs. Every NEP proposal I'm familiar with makes use of some form of space-going reactor (i.e. energy production by fission rather radioactive decay), such as the Russian Topaz reactor (5+ kW electrical output).
Objects look the way they do because of where they are in relation to the viewer and what their dimensions are. Similarly light has rules which you can learn if you are to duplicate the illusion of light in a 2d representation like a drawing. Textures also have rules and so on. It's all about drawing what you see and not what you think is there.
Which is fine for photo-realism, but won't produce original art. Admittedly, you can get tips on things like composition, color choice, and so on too. But still, it's more art than science. That's what makes it art. It's mostly about practice, along with a little native talent.
Hey, I'm not saying that there aren't hypocritical idiots in the "free-marketer" camp. Just that not everyone in that camp is an idiot:)
Kind of depends on whether you support the "free market" because you are truly interested in freedom and voluntary association, or if you are just using it as a code-word for "let corporations do whatever they like" (while conveniently forgetting that corporations are themselves an artifact of government interference in the market).
I suspect that (speaking as a bit of a "free-marketer" myself) the complaints arise when unions attempt to use the force of government to make companies use union labor no matter what, or attempt to obtain monopoly status for their union in a particular industry (union-only shops). I don't mind unions that are voluntary to join - if the union provides reasonable benefits then they'll get lots of members.
Slashdot Mirror
Well, Debian for one. In fact, Debian supports x86, Motorola 68k, SPARC, Alpha, PowerPC, ARM, MIPS, PA-RISC, IA-64, and S/390 architectures. Porting to the AMD64 and Hitachi SuperH is also underway. Note that the 68k, PowerPC, ARM, and SuperH are all popular for embedded applications.
Ok, seriously, are you just trolling here? Or are you really trying to equate a loss of a life-saving antibiotic with the "life and death" prospect of missing the next episode of 'Survivor'?
Uh... it's just TV... you don't have to watch it you know.
Well, there is a CSP/Occam-derived add-on for Java that supports a robust concurrency model. It just isn't widely used (yet). Note that there are similar concurrency toolkits for C and C++ (in fact I believe that the x86 port of Occam is built on top of the CCSP toolkit).
The sad thing is that the Transputer is still a great idea - it died mostly due to poor business decisions by INMOS (late getting newer processors out). I mean, look at the popularity of SMP, Beowulf clusters (that perenial /. favorite), and the like. The transputer was specifically designed to support that kind of thing in a clean way, and had a language (occam) that made it easy to program for multiprocessor systems. Sigh. Sometimes it seems like the computer industry has a habit of ignoring great innovations, and then reinventing them badly a few years later.
It's probably worth adding that from the TFA it seems that VxWorks code is shared between different spacecraft programs. So the folks who can test on real satellites are sharing their patches, fixes, and features.
There are several open source RTOS's out there, if you want to provide some kind of educational aid. Off the top of my head, I can think of RTEMS, eCos, and RT-Linux. I've seen several real-time courses use the MicroC OS, but I don't recall if it is open source or not. The odds of WindRiver (NASA doesn't own the code) open sourcing VxWorks are pretty minimal I would imagine.
Shielding does not protect against single-event upsets (particle-induced bit flips), it only provides some mitigation against total ionizing dose (which causes long term cumulative degradation as a result of drift in transistor operating parameters). There are design techniques and fabrication processes that can reduce the likelihood that a circuit will suffer upsets, but it's still standard practice to provide either redundant memory, or error detection and correction coding. In the case of MER they had 3 physically separate PROMs carrying identical copies of the flight software, and the RAM was (IIRC) protected by an EDAC code implemented in a rad-hard FPGA.
You are confusing Mars Pathfinder (a 1997 mission, which suffered a priority inversion problem) with Mars Exploration Rover (a 2003-2004 mission which suffered a file allocation issue in Flash memory, and the subject of TFA). Although both used VxWorks.
Having said all of that, I'll agree that formal verification at NASA is in its infancy, and is facing an uphill battle for acceptance (witness how long the Langley group has been trying to push formal methods). It'll be interesting to see what happens with JPL's LaRS.
Or perhaps because NASA doesn't own the code -WindRiver does.
Actually, there are some spectacularly rich people in rural areas. They just aren't all that showy about it. Similarly, many of the folks in LA that drive SUVs are living beyond their means, but obsessed with image.
Actually, I saw far more SUVs when I was living in LA than I do now that I'm living in a "Red State".
Considering that the Chinese government builds and operates its own launch vehicles I doubt that launch costs are an issue. That said, your point about spatial resolution is bang on, and is the number one reason I can't imagine that they'd mess with a GEO "spy sat".
Spy sat fly low because the required camera resolution and spacecraft pointing accuracy is far lower. They fly in highly inclined orbits because that allows them to cover the most latitude - polar orbits will cover the entire world. They also fly highly-inclined because it allows them to achieve a sun-synchronous orbit that provides good lighting conditions for taking imagery.
And actually, the original poster was correct in his usage - geosynchronous simply means a 24 hour circular orbit (35,786 km altitude). Geostationary refers to a geosynchronous orbit with a (nominally - there is some drift) 0 degree inclination, which minimizes the amount of apparent north-south motion and keeps the spacecraft at what appears to be a fixed point in the sky.
The problem is that spacecraft power consumers consist of more than just electronics. And yes, electronics have improved since the 70's, but they also do a lot more than their 70's ancestors do. You will consume 60-70 W in comm system alone (transponders plus amplifiers), even neglecting high-rate transmissions. Your onboard computer (plus associated interface circuits) will consume upwards of 30 W (for a non-redundant system - more for a redundant system). Survival heaters will probably consume anywhere from 70 W up to 200 W (yes, you can use some isotope-based heaters, but they cannot be switched on and off, so they are of limited use). Between reaction wheels, star trackers, IMUs, and the like, your attitude control system will consume 70+ W at a minimum. None of this includes the inevitable inefficiencies in DC-DC conversion and power distribution. These are real estimates, based on real deep-space (as opposed to near-earth) missions (NEAR, DAWN, ACE, etc).
"Non-uniform" gravitation has little or nothing to do with it. The prinicipal disturbance torques acting on a spacecraft are gravity-gradient (which has little to do with the non-uniformity of the field), magnetic, solar radiation, and aerodynamic torques. Aerodynamic torques are negligible even for high LEO orbits. GG torques will not affect a deep sapce mission. The other two are still a concern. Plus, you are constantly turning your spacecraft to keep the thrust vector pointed in the appropriate direction.
Not really. Why would you design a craft to constantly be expecting to receive when you have a delay time of hours? Scheduled command receipt times make far more sense, unless you've just got power to burn.
[sigh] Look, no one (or at least no one I am aware of) designs spacecraft that switch their command receiver on and off - there's too much chance that it won't get switched on again at the right time (or ever). Plus, if you are constantly thrusting you are going to want to get regular telemetry, just to amke sure everything's still operating correctly.
The fact remains that with 1970s tech, the Voyager craft idled on minimal power. Now you're trying to say that it's not realistic with 2000s tech. That rings incredibly hollow.
See above - we're talking at least 200 W, probably more. And these are not subsystem you can switch off: if you are constantly thrusting you need to be controlling your attitude - which means ACS on, flight computer on, comm system on for telemetry and command updates. You can't "idle" at 0 W, or even close to it.
If it's not enough for you, then we can upgrade the power supply; Cassini, after all, launched with ~750W of RTG power (of which a peak load is only needed during flybys) - and there, we're not even talking about a craft that powered its propulsion via electricity. If you want to devote even a small fraction of that saved propellant mass to RTG power, you can get a T-140 without having to launch an unreasonable amount of plutonium (i.e., the costs of the plutonium will still be insignificant compared to the costs of the entire project). Then, with your remaining propellant savings, you can fill it with insturments, and you can have your insturments make use of your higher power supply as well. It's a win-win situation.
The T-140 requires a minimum of 1.8 kW. If you're going to try to supply that much power (plus whatever you need for spacecraft power), why not just use a reactor? You are now attempting to fly almost of 500 kg of RTGs, versus a reactor that could weigh less than 500 kg, and produce significantly more power output (while using cheaper uranium, instead of plutonium). That's been my point al
Voyagers 1 and 2 were deliberately built simple and rugged, and were tightly constrained by the power available from their RTGs. Modern spacecraft are more complex and often consume more energy. Spectrum Astro's DS1 factsheet gives the bus orbit average power as 500 W. Admittedly, that probably isn't with all systems "idled" (whatever that means), but there is a limit to how much you can switch off on a spacecraft in mid-flight.
I can run a graphing calculator for a day on a couple double a batteries that has 10 times the processing power of these craft
A graphing calculator is not an internally redundant rad-hard flight computer (30+ W), with its associated I/O interfaces. It does not have a comm system (which will draw power even when not transmitting - you always want to be able to receive). It does not have attitude sensors. It does not have attitude actuators (which can be large power consumers on a 3-axis stabilized spacecraft, as you would want to be with a low-thrust system since spin-axis precession is probably infeasible). These things all take power. 500 W? Not necessarily. Several hundred W. Quite probably.
That the cost of the plutonium is completely dwarfed by the cost of launching the fuel?
The cost is not just the cost of plutonium itself, but the additional cost and complexity that results from flying a radioactive (and toxic) substance. Cassini cost $1.2 billion just to develop - the launch cost is a small fraction of the development cost, let alone the total mission cost. Admittedly, not all of Cassini's cost was driven by the RTGs, but they definitely contribute to a higher cost than you would otherwise see.
Hall effect thrusters, while having ISPs of "only" around 2,500 (compared to ~450 for a good LOX/LH engine) give power/thrust ratios of near .65-.70 newton per kwH in the case of the T-220 (assumedly the T-40 performs similarly). A 200W T-40 HET, therefore, would well outperform DS1's engine operating at full power, and yet use a tiny fraction of the propellant of a chemical engine.
Unfortunately, electric propulsion system performance doesn't tend to scale linearly. There are economies of scale that come with larger systems. Quoting from the Pratt & Whitney website:
a single Cassini RTG could power a T-40 HET with power to spare (you can bring more RTGs as you want for insturments) and produce over 100 millinewtons of continuous thrust. Over a year, it could impart, to a 1000 kg spacecraft, over 3000 m/s delta-V**. Seing as it took Gallileo 5 years to get to Jupiter, that's 15k m/s delta-V** - I.e., you could get to Saturn using *no* gravity assists as fast as Gallileo did *using* gravity assists.
Since you have overestimated HET performance you are getting numbers that are unrealistic (which is exacerbated by the exponential dependence of propellant mass on Isp). Using your numbers (which are optimistic) we get:
However, if we use realistic (but still optimistic - I'm assuming the upper end of the quoted T-40 performance range, which requires closer to 400 W) numbers for low-power HET performance, we get**
From a NASA report describing their Stirling Cycle radioisotope generator - real efficiencies, not theoretical numbers.
Also, RTGs don't use Peltier-type devices (thermoelectrics) - they're actually thermionic. Thermoelectric ("Peltier-type") devices use a junction in electrically conductive/heat insulating materials. Thermionics use a vacuum between two extreme-heat-differential plates, where the hot-plate acts like a cathode and emits electrons.
I don't know where you are getting you're information from. There may be a few "RTGs" that use thermionics for energy conversion (I haven't heard of any though). The majority of RTGs do use thermoelectrics. For example, Cassini's RTGs make use of SiGe unicouples for energy conversion. Thermionics are much more common in space-going reactors, such as Topaz, Topaz II, and the US SP-100.
In-transit? You don't need much insturment power for the dozen or so years to the outer planets. What, do you think they're gathering nonstop radar images of deep space while flying?
It's not instrument power that's the issue. It's power for the comm system, the onboard computers, the attitude control system, and the thermal control system. Yes, you can reduce the amount of pwoer that these things draw, but you cannot make it zero, and we are still talking about a significant number of Watts.
As I mentioned, DS1 *did* use just a couple hundred watts for its ion engine for a good portion of its flight, so it clearly is a reasonable option. Hall effect thrusters use even less electricity for the same total amount of thrust, with a penalty of lower ISP (but still far greater than chemical rockets).
And as I mentioned, DS1 had a power source that produced far more than the amount of energy that it needed for the thruster - running both spacecraft systems and thruster off of "a few hundred watts" is not practical. The DS1 spacecraft, excluding the thruster, used about 400-500 W of power (see my points above about other subsystems that require power).
Why exactly would you not? Ion thrusters have plenty of time (months to years) to line up the craft for planetary assists. You might have trouble getting assists from moons when in orbit around a planet due to the short differences in time between encounters, but that's about it.
There is plenty of work being done on combing low-thrust trajectories and gravity-assist trajectories. I don't dispute that there is a lot of potential there. The question is whether or not the RTG-based design that you are proposing would be worthwhile in such a mission. I am saying no. See below.
Given standard launch cost rates (~25k$/kg for escape velocity), it would cost about 140 million dollars to launch Cassini. 78 million of that would go to launching just its propellant. RTG costs are trivial in comparison to propellant launch costs.
Actually, it cost closer to $450 million to launch Cassini, since they used an ultra-expensive Titan LV. The problem with your math is that launches aren't sold by the kg, they're sold by the launch vehicle. Would Cassini have been light enough to get to a smaller launch vehicle if you switched to ion thrusters? Hard to say - yes, you will get some propellant mass savings due to better Isp. But, either you use the existing power system, which means very low thrust, a seriously extended mission duration, and a greater need for redundancy, or you beef up the power system (use something other than RTGs). The latter approach being the one the JIMO has taken.
10x the ISP is *NOT* minor gains. 10x the ISP is a huge gain.
Yes, it is. But you are failing to account for the additional mass of the power processing infrastructure that is necessary to support an ion thruster, or the extra thermal radiators you will need due to the inefficiencies of
JIMO has, AFAIK, always been baselined to use a reactor. It's the only thing that makes sense. Particularly since they are talking in terms of hundreds of kW of power.
Apart from the fact that NASA is developing higher power stirling cycle RTGs
If it's Stirling Cycle it's not and RTG - RTG stands for Radioisotope Thermoelectric Generator, and uses Peltier-type devices to directly convert thermal energy into electrical energy. The drawback is that the conversion process is not very efficient (although it does avoid the need for moving parts). The reason that NASA is looking into Stirling Cycle Radioisotope Generators is that they can boost the energy conversion efficiency to maybe 20%. But you still only get a relativle small amount of pwoer out of a Radioisotope Generator of this kind (I believe that they are talking about 55 We from ~500Wth). AFAIK Europa Orbiter was a NASA (not ESA) mission that used chemical (not electric) propulsion, so it's not really relevant to this discussion.
Deep Space 1, for example, operated its ion drive for months at 520 watts
That's nice, but what are you supposed to run the rest of the spacecraft on? DS1 had solar arrays that produced ~2.5 kW at BOL, so even at max power on the ion thruster they still had a bunch of power left over to run the other spacecraft systems. I won't dispute that HET's are somewhat more power efficient, but not enough to make a significant difference compared to electrostatic ion thrusters.
I'm not arguing that low-power ion thrusters powered by an RTG won't produce more thrust in a Jovian orbit that a SEP system (although it depends a lot on the size of your solar arrays). I'm arguing that the thrust levels that you can achieve using any power source that produces only a few hundred Watts for both spacecraft and thruster is insufficient to perform a deep-space mission - the resulting trip times become unimaginably long. Now, perhaps if you were to combine traditional ballistic gravity-assists with RTG-based propulsion for say minor stationkeeping maneuvers you might be on to something. But somehow I doubt the (high!) cost and complexity of using an RTG would be worth the minor gains that you might make in propellant savings.
There is a lot of interest in so-called "Nuclear-Electric Propulsion" (NEP) right now, but no one is talking RTGs. Every NEP proposal I'm familiar with makes use of some form of space-going reactor (i.e. energy production by fission rather radioactive decay), such as the Russian Topaz reactor (5+ kW electrical output).
Which is fine for photo-realism, but won't produce original art. Admittedly, you can get tips on things like composition, color choice, and so on too. But still, it's more art than science. That's what makes it art. It's mostly about practice, along with a little native talent.
Kind of depends on whether you support the "free market" because you are truly interested in freedom and voluntary association, or if you are just using it as a code-word for "let corporations do whatever they like" (while conveniently forgetting that corporations are themselves an artifact of government interference in the market).
I suspect that (speaking as a bit of a "free-marketer" myself) the complaints arise when unions attempt to use the force of government to make companies use union labor no matter what, or attempt to obtain monopoly status for their union in a particular industry (union-only shops). I don't mind unions that are voluntary to join - if the union provides reasonable benefits then they'll get lots of members.