Unfortunately the math doesn't work out for trips from Earth to Mars. A laser, like any other source of light, is still subject to the 1/r^2 law. The r is the effective radius of the lens, which you can make pretty big, but Earth to Mars is an enormous distance and only a tiny fraction of the energy would hit a solar sail a few kilometers across with a practical mirror you could make in orbit.
It's ok to be mad, but keep in mind that the engineers of an entertainment service deliberately did not engineer it to be five 9s reliable BECAUSE it's entertainment. They knew they could save a tremendous amount of money by designing it for a lower tier of reliability. Also, you paid only $9 a month, less than the cost of a plain old phone line(which is that reliable), for streaming HD video of thousands of titles AND the production of new content. As long as it only goes down for a few hours a year, you got one hell of a deal.
Frankly, you didn't pay for a mission-critical reliable service, so quit whining.
What if it comes with a "look but no exact copies license". (where an exact copy is defined as a product that does not have at least 20% more or different parts)?
What if it comes with a "look but any clones have to be sold for a minimum price" license?
I hate to defend Trump here but there are ways you can screw it up that are understandable. Trump didn't know how popular his casino would be. There was an overall market shift happening where the market for casinos has become saturated.
Umm, did the upper group gain income faster than their tax burden increased?
The lower group may have had a reduction in taxes paid, but the lower group also faces gigantically inflated prices for medical care and housing. The upper group also faces said increased cost, but obviously it's a vastly smaller proportion of their income. So overall, whether or not the taxes shifted, most Americans (the lower group is the overwhelming majority of all the people) have done worse.
More or less. Uncrewed nuclear-electric "barges" have the obvious problem that they are technically very complex and if anything goes wrong, there is no crew aboard able to conduct repairs. Also, nuclear reactors meant for space might be extremely expensive, the cost exceeding the cost of a few heavy-lift rocket runs full of fuel. As Mr. Musk has demonstrated, it's entirely possible to launch rockets that are mostly reusable, so the cost of a few swimming pools full of additional propellant might not be as expensive as you would expect.
You might be able to solve the maintenance problem with semi-autonomous robots that can carry out tasks and even learn, told in general terms what to do from a control station on earth.
Yes but the heavier your power generator needs to be. So the higher the ISP, the worse the effective thrust. The worse the thrust, the longer it takes to get somewhere.
You start the Mars injection burn in high orbit. You end it on a trajectory that will forever remain in solar orbit if no more engine burns are made. The engine is not radioactive when it is put in a rocket and launched, in fact you launch the engine unfueled and launch the fuel sections in separate rockets. Each fuel section would not have enough fuel to form a critical mass, so even in the worst possible accident the damage would be limited.
It's a cylinder. Inside it are rods. The rods have channels in between. There is HEU fuel inside the rods and they are made of high temperature alloys.
For a practical Mars mission, your spacecraft would have 2-4 NERVA engines, each a little smaller than the last. On earth departure, you eject the rods that block the fuel channels and dampen the reaction. Engine is cold and has never been run at that point. You run the engine until the burn is complete. Then separate the hot radioactive mess of an engine that is melting down because when the propellant flow stops, so does the cooling.
That's the simplest way to do it. No nuclear power reactor, it's 1 time use. The pumps, radiation shield, and so forth you don't stage separate, you just eject the core and pressure vessel cannister it is. (so it can be extremely light, under a metric ton)
Ok, so 3 kg / kw electric is 3 times worse than it "needs to be" for a 39 day trip to Mars.
Terrible. And if I actually look at the chart you linked on page 3, that's solid core fusion (0.15 kg/kWout) + Rankine Heat Engine(.14/kg heat IN) + Neutron Shield(1 per kw HEAT) + single phase radiator, 600 K(.98 kg per kilowatt HEAT).
I already know this is going to be terrible? Why? Because it's talking about HEAT kilowatts. With 600 K radiator (efficiency is even shittier at higher radiator temperatures, see the laws of thermodynamics for a source), you get maybe 33% efficiency with the Rankine cycle. I can work out the exact numbers if you demand but it's not going to matter.
so 3 * 0.15 + 3* 0.14 + 3 * 1 + 3*.98.
6.84 kilograms of heavy ass shit per kilowatt of electric power supply, best case, using conventional nuclear generators. Zubrin was right. You were talking out your rectum. Disagree? Prove it with numbers.
Sure, you can do a shit ton better with Aneutronic fusion and TWDEC. Direct conversion is 90% efficient so your heat to reject just plummeted. Go ahead and use a droplet radiator at 1000 K. And use the electricity from direct conversion to run the magnets of _another_ fusion engine where it crushes pellets of fusion fuel and the products of the reaction all escape at high velocities out the exhaust. Now we're talking, an engine that gives a respectable ISP and at the same time is lightweight so (relatively) high thrust. But that's far, far future. Scientists haven't even managed to properly contain fusion plasma yet and have only the vaguest napkin sketches of such a device.
1 kw/kg is still a tall order. You end up needing to use things like droplet radiators to even approach a number like that. As it so happens, solar panels are inherently much lighter because they can be a paper thin sheet of layered junctions - you can actually get much higher power/mass with solar than any heat engine nuclear reactor design anyone has drawn up. The only reason it sucks to use a solar-electric spacecraft to go to Mars is the sunlight incidence per square meter plummets to 44% out there.
In any case, Zubrin is basically correct. You may save on radiation shielding and propellant mass even if you _could_ make a nuclear electric thruster work (and you don't need vasimr - a big array of Hall or MPD or 30 other ways would work) but now you have to launch and run in flight that nasty high power nuclear reactor. He's totally right about that in that you're trading off a simple system (some tanks of propellant and/or a simple nuclear thermal rocket) for whatever exotic system you need for that level of power density.
He wants to see someone make it to Mars before he personally dies of old age, and nuclear-electric is not the way to accomplish that. This is understandable.
Oh, since you probably won't read the whole article, the key paragraph is this :
"But wait, there’s more. To achieve his much-repeated claim that VASIMR could enable a 39-day one-way transit to Mars, Chang Diaz posits a nuclear reactor system with a power of 200,000 kilowatts and a power-to-mass ratio of 1,000 watts per kilogram. In fact, the largest space nuclear reactor ever built, the Soviet Topaz, had a power of 10 kilowatts and a power-to-mass ratio of 10 watts per kilogram. There is thus no basis whatsoever for believing in the feasibility of Chang Diaz’s fantasy power system."
What is key about this paragraph is the argument applies to _any_ electric thruster, not just vasimr. All are basically useless for going to Mars.
No it isn't. TLDR, read atomic rockets, n00b. Or, since you won't be reading it : high specific impulse ion drives have existed in various test forms for years. They are easy to make and they all have high efficiency, albeit some designs are more reliable than others. The problem with all of them is the nasty equation KE = 1/2 m * V^2. That means the higher the exhaust velocity (and thus specific impulse which is the fuel efficiency), energy required goes up with the square of exhaust velocity.
Plenty of ion thruster designs, including VASIMR, have reasonable energy efficiency. The problem is that you still have to pay the bill even if the efficiency were 100%. You still have to supply as much electrical power as the kinetic energy of the escaping propellant.
This is a big problem. Even the most exotic nuclear power generator designs anyone has drawn up, the nuclear generator is a heavy ass piece of equipment being propelled by barely any resulting thrust from the ion drive. It means that you might have great specific impulse but trips to Mars still take months. What you need is also high thrust. That's why the best engine for space travel that currently is feasible is still plain old nuclear-thermal. You only get an ISP of about 1000 (compare to 15k for this particular thruster), but you get thousands of times more thrust. You can complete your Mars injection burn in about half an hour instead of having to run the engine for months. That in turn increases efficiency because there is always 1 optimal point to do your engine burns at.
Yep. I turned them off because I just don't really see the details of other people's cars when I'm driving. At most I just think of them as car/truck/18 wheeler, and as obstacles to be avoided. So there could be "silver cadillac plate number XXX-XXX" as an alert, and I could be right behind it on the highway for an hour 15 minutes after receiving the alert, and I just wouldn't see it.
You know...amber alerts + autonomous cars would actually work pretty damn well. Machine vision, of the quality needed to drive an autonomous car, is more than good enough to read the license plate and to identify the color and probably usually the make/model of a given car your autonomous car can see. If the alerts went right to the computers of autonomous cars, and autonomous cars who think they detected the matching vehicle sent automatic texts with the photograph of the vehicle back to the government, it would actually work. I mean, other than the dystopian possibilities, ofc.
But PMSM motors are better. And you can design the motor to be more efficient at low speeds. Also there's gearing, though you want to use fixed gear ratios obviously to avoid the mechanical complexity of a transmission, there's electrical gearing as an option. Even 2 parallel sets of windings in the motor, around a common shaft.
Obviously, from the perspective of motor windings, a 2 pole motor at 100 rpm is the same as a 4 pole motor at 50 rpm.
Ok, I googled for it, and it seems that the solution is more complex than just using an inverter : https://motorino.ca/technology...
The point is, it can be done. It costs more in components than the cheapest possible scooter (but it's still well within the capabilities of Chinese mass industry to manufacture) but it would give you enough power that hills wouldn't be a problem. All is required is the government gradually phase out the gasoline scooters by increasing taxes and fees, slowly ratcheting up year after year until the annual registration fees are more than cost difference to electric. Then just make them illegal.
Not true. In fact, with AC power supplies, you go AC->DC->inverter FETs. DC->inverter FETs is less parts. You clearly were talking out your ass it seems. This would be a matter of motor design - I'm not going to say for certain if you can design a motor to give you high enough torque at low speeds and also have high efficiency at high speeds, but you probably can.
Perhaps. The thing is, there are people who have been programming in C since they were 11, and they are 30 now. Basically no one has that depth of experience in a rare language like Erlang.
I mainly do embedded systems work where C with inline ASM is the only viable option, but I do wish for something better than C++ when it comes time to do the pieces of the system that are higher level. C++ is a bloated mess and has many design faults that make catastrophic bugs far more likely. C is simple and fast but lacks many of the abstractions needed to build a complex system efficiently.
But what can I do. Qt is a great platform for embedded guis, but it is pure C++. It's one of the most _supported_ platforms, you can get embedded processor OS images that are already set up to go straight to running your GUI in Qt with the least hassle.
I've looked at a little hobbyist game development - UE4 is rapidly becoming the go to engine for that, and it's almost entirely C++.
So even though I do wish for a language that doesn't have the drawbacks of C++, all I can really do is read Bjarne Stroustrup's books and learn the lesser used features so I have a chance in interpreting other people's confusing code...(with so many language features, you can easily create code that other people won't know how to read)
1. More popular languages you can find less expensive developers for...(or more total developer talent for a given amount of money)
2. More popular languages are likely to have stackoverflow posts with examples for every single basic task and explanations about what errors mean. (before you gloat about how you don't need help from SO, have you ever faced goddamn C++ linker errors? Might as well have an Ouija board out to find out what the mistake is)
3. More popular languages work on more computers and are likely to continue to work in the future
4. More popular languages tend to be faster. Usually a shit ton faster. Java has gone from a bloated mess to a bloated mess that is often within spitting distance of C on performance shootouts. That's from the popularity spurring further development. C is almost always king of the hill and nothing is faster. Python? Rust? Whatever n00bs. Those languages may be nice to write complex code that only gets run occasionally but if you need high end performance they aren't going to cut it.
Slashdot Mirror
Unfortunately the math doesn't work out for trips from Earth to Mars. A laser, like any other source of light, is still subject to the 1/r^2 law. The r is the effective radius of the lens, which you can make pretty big, but Earth to Mars is an enormous distance and only a tiny fraction of the energy would hit a solar sail a few kilometers across with a practical mirror you could make in orbit.
Because nothing else on the table exists as a real system.
It's ok to be mad, but keep in mind that the engineers of an entertainment service deliberately did not engineer it to be five 9s reliable BECAUSE it's entertainment. They knew they could save a tremendous amount of money by designing it for a lower tier of reliability. Also, you paid only $9 a month, less than the cost of a plain old phone line(which is that reliable), for streaming HD video of thousands of titles AND the production of new content. As long as it only goes down for a few hours a year, you got one hell of a deal.
Frankly, you didn't pay for a mission-critical reliable service, so quit whining.
What if it comes with a "look but no exact copies license". (where an exact copy is defined as a product that does not have at least 20% more or different parts)?
What if it comes with a "look but any clones have to be sold for a minimum price" license?
I hate to defend Trump here but there are ways you can screw it up that are understandable. Trump didn't know how popular his casino would be. There was an overall market shift happening where the market for casinos has become saturated.
Umm, did the upper group gain income faster than their tax burden increased?
The lower group may have had a reduction in taxes paid, but the lower group also faces gigantically inflated prices for medical care and housing. The upper group also faces said increased cost, but obviously it's a vastly smaller proportion of their income. So overall, whether or not the taxes shifted, most Americans (the lower group is the overwhelming majority of all the people) have done worse.
More or less. Uncrewed nuclear-electric "barges" have the obvious problem that they are technically very complex and if anything goes wrong, there is no crew aboard able to conduct repairs. Also, nuclear reactors meant for space might be extremely expensive, the cost exceeding the cost of a few heavy-lift rocket runs full of fuel. As Mr. Musk has demonstrated, it's entirely possible to launch rockets that are mostly reusable, so the cost of a few swimming pools full of additional propellant might not be as expensive as you would expect.
You might be able to solve the maintenance problem with semi-autonomous robots that can carry out tasks and even learn, told in general terms what to do from a control station on earth.
That's what I said. _Although_ you could use droplet radiators to reduce that problem hugely.
Yes but the heavier your power generator needs to be. So the higher the ISP, the worse the effective thrust. The worse the thrust, the longer it takes to get somewhere.
You start the Mars injection burn in high orbit. You end it on a trajectory that will forever remain in solar orbit if no more engine burns are made. The engine is not radioactive when it is put in a rocket and launched, in fact you launch the engine unfueled and launch the fuel sections in separate rockets. Each fuel section would not have enough fuel to form a critical mass, so even in the worst possible accident the damage would be limited.
It's a cylinder. Inside it are rods. The rods have channels in between. There is HEU fuel inside the rods and they are made of high temperature alloys.
For a practical Mars mission, your spacecraft would have 2-4 NERVA engines, each a little smaller than the last. On earth departure, you eject the rods that block the fuel channels and dampen the reaction. Engine is cold and has never been run at that point. You run the engine until the burn is complete. Then separate the hot radioactive mess of an engine that is melting down because when the propellant flow stops, so does the cooling.
That's the simplest way to do it. No nuclear power reactor, it's 1 time use. The pumps, radiation shield, and so forth you don't stage separate, you just eject the core and pressure vessel cannister it is. (so it can be extremely light, under a metric ton)
Ok, so 3 kg / kw electric is 3 times worse than it "needs to be" for a 39 day trip to Mars.
Terrible. And if I actually look at the chart you linked on page 3, that's solid core fusion (0.15 kg/kWout) + Rankine Heat Engine(.14/kg heat IN) + Neutron Shield(1 per kw HEAT) + single phase radiator, 600 K(.98 kg per kilowatt HEAT).
I already know this is going to be terrible? Why? Because it's talking about HEAT kilowatts. With 600 K radiator (efficiency is even shittier at higher radiator temperatures, see the laws of thermodynamics for a source), you get maybe 33% efficiency with the Rankine cycle. I can work out the exact numbers if you demand but it's not going to matter.
so 3 * 0.15 + 3* 0.14 + 3 * 1 + 3*.98.
6.84 kilograms of heavy ass shit per kilowatt of electric power supply, best case, using conventional nuclear generators. Zubrin was right. You were talking out your rectum. Disagree? Prove it with numbers.
Sure, you can do a shit ton better with Aneutronic fusion and TWDEC. Direct conversion is 90% efficient so your heat to reject just plummeted. Go ahead and use a droplet radiator at 1000 K. And use the electricity from direct conversion to run the magnets of _another_ fusion engine where it crushes pellets of fusion fuel and the products of the reaction all escape at high velocities out the exhaust. Now we're talking, an engine that gives a respectable ISP and at the same time is lightweight so (relatively) high thrust. But that's far, far future. Scientists haven't even managed to properly contain fusion plasma yet and have only the vaguest napkin sketches of such a device.
Figures of 1000W/kg aren't just some sort of out-of-a-hat fantasy numbers, they're based on the very real work in the field.
Care to cite? I've looked fairly extensively for anything on this, and no, everything I read said it was fantasy.
1 kw/kg is still a tall order. You end up needing to use things like droplet radiators to even approach a number like that. As it so happens, solar panels are inherently much lighter because they can be a paper thin sheet of layered junctions - you can actually get much higher power/mass with solar than any heat engine nuclear reactor design anyone has drawn up. The only reason it sucks to use a solar-electric spacecraft to go to Mars is the sunlight incidence per square meter plummets to 44% out there.
In any case, Zubrin is basically correct. You may save on radiation shielding and propellant mass even if you _could_ make a nuclear electric thruster work (and you don't need vasimr - a big array of Hall or MPD or 30 other ways would work) but now you have to launch and run in flight that nasty high power nuclear reactor. He's totally right about that in that you're trading off a simple system (some tanks of propellant and/or a simple nuclear thermal rocket) for whatever exotic system you need for that level of power density.
He wants to see someone make it to Mars before he personally dies of old age, and nuclear-electric is not the way to accomplish that. This is understandable.
Oh, since you probably won't read the whole article, the key paragraph is this :
"But wait, there’s more. To achieve his much-repeated claim that VASIMR could enable a 39-day one-way transit to Mars, Chang Diaz posits a nuclear reactor system with a power of 200,000 kilowatts and a power-to-mass ratio of 1,000 watts per kilogram. In fact, the largest space nuclear reactor ever built, the Soviet Topaz, had a power of 10 kilowatts and a power-to-mass ratio of 10 watts per kilogram. There is thus no basis whatsoever for believing in the feasibility of Chang Diaz’s fantasy power system."
What is key about this paragraph is the argument applies to _any_ electric thruster, not just vasimr. All are basically useless for going to Mars.
Yes. http://spacenews.com/vasimr-ho...
No it isn't. TLDR, read atomic rockets, n00b. Or, since you won't be reading it : high specific impulse ion drives have existed in various test forms for years. They are easy to make and they all have high efficiency, albeit some designs are more reliable than others. The problem with all of them is the nasty equation KE = 1/2 m * V^2. That means the higher the exhaust velocity (and thus specific impulse which is the fuel efficiency), energy required goes up with the square of exhaust velocity.
Plenty of ion thruster designs, including VASIMR, have reasonable energy efficiency. The problem is that you still have to pay the bill even if the efficiency were 100%. You still have to supply as much electrical power as the kinetic energy of the escaping propellant.
This is a big problem. Even the most exotic nuclear power generator designs anyone has drawn up, the nuclear generator is a heavy ass piece of equipment being propelled by barely any resulting thrust from the ion drive. It means that you might have great specific impulse but trips to Mars still take months. What you need is also high thrust. That's why the best engine for space travel that currently is feasible is still plain old nuclear-thermal. You only get an ISP of about 1000 (compare to 15k for this particular thruster), but you get thousands of times more thrust. You can complete your Mars injection burn in about half an hour instead of having to run the engine for months. That in turn increases efficiency because there is always 1 optimal point to do your engine burns at.
Yep. I turned them off because I just don't really see the details of other people's cars when I'm driving. At most I just think of them as car/truck/18 wheeler, and as obstacles to be avoided. So there could be "silver cadillac plate number XXX-XXX" as an alert, and I could be right behind it on the highway for an hour 15 minutes after receiving the alert, and I just wouldn't see it.
You know...amber alerts + autonomous cars would actually work pretty damn well. Machine vision, of the quality needed to drive an autonomous car, is more than good enough to read the license plate and to identify the color and probably usually the make/model of a given car your autonomous car can see. If the alerts went right to the computers of autonomous cars, and autonomous cars who think they detected the matching vehicle sent automatic texts with the photograph of the vehicle back to the government, it would actually work. I mean, other than the dystopian possibilities, ofc.
But PMSM motors are better. And you can design the motor to be more efficient at low speeds. Also there's gearing, though you want to use fixed gear ratios obviously to avoid the mechanical complexity of a transmission, there's electrical gearing as an option. Even 2 parallel sets of windings in the motor, around a common shaft.
Obviously, from the perspective of motor windings, a 2 pole motor at 100 rpm is the same as a 4 pole motor at 50 rpm.
Ok, I googled for it, and it seems that the solution is more complex than just using an inverter : https://motorino.ca/technology...
The point is, it can be done. It costs more in components than the cheapest possible scooter (but it's still well within the capabilities of Chinese mass industry to manufacture) but it would give you enough power that hills wouldn't be a problem. All is required is the government gradually phase out the gasoline scooters by increasing taxes and fees, slowly ratcheting up year after year until the annual registration fees are more than cost difference to electric. Then just make them illegal.
Not true. In fact, with AC power supplies, you go AC->DC->inverter FETs. DC->inverter FETs is less parts. You clearly were talking out your ass it seems. This would be a matter of motor design - I'm not going to say for certain if you can design a motor to give you high enough torque at low speeds and also have high efficiency at high speeds, but you probably can.
They are? Even inverter drive motors?
Perhaps. The thing is, there are people who have been programming in C since they were 11, and they are 30 now. Basically no one has that depth of experience in a rare language like Erlang.
I mainly do embedded systems work where C with inline ASM is the only viable option, but I do wish for something better than C++ when it comes time to do the pieces of the system that are higher level. C++ is a bloated mess and has many design faults that make catastrophic bugs far more likely. C is simple and fast but lacks many of the abstractions needed to build a complex system efficiently.
But what can I do. Qt is a great platform for embedded guis, but it is pure C++. It's one of the most _supported_ platforms, you can get embedded processor OS images that are already set up to go straight to running your GUI in Qt with the least hassle.
I've looked at a little hobbyist game development - UE4 is rapidly becoming the go to engine for that, and it's almost entirely C++.
So even though I do wish for a language that doesn't have the drawbacks of C++, all I can really do is read Bjarne Stroustrup's books and learn the lesser used features so I have a chance in interpreting other people's confusing code...(with so many language features, you can easily create code that other people won't know how to read)
But it isn't as fast as C. Look at the benchmarks.
Of course popularity matters.
1. More popular languages you can find less expensive developers for...(or more total developer talent for a given amount of money)
2. More popular languages are likely to have stackoverflow posts with examples for every single basic task and explanations about what errors mean. (before you gloat about how you don't need help from SO, have you ever faced goddamn C++ linker errors? Might as well have an Ouija board out to find out what the mistake is)
3. More popular languages work on more computers and are likely to continue to work in the future
4. More popular languages tend to be faster. Usually a shit ton faster. Java has gone from a bloated mess to a bloated mess that is often within spitting distance of C on performance shootouts. That's from the popularity spurring further development. C is almost always king of the hill and nothing is faster. Python? Rust? Whatever n00bs. Those languages may be nice to write complex code that only gets run occasionally but if you need high end performance they aren't going to cut it.