It’s been a long time since I read The Fountains of Paradise, and I don’t have a copy, but I’m pretty sure that Clarke has an astronaut die during construction of one of the intermediate waystations below geosynchronous orbit - betrayed by years of experience in orbital construction work, he stepped off the platform untethered and fell. It could have been at the counterweight and he was flung.
The only death I recall is the protagonist while rescuing people stranded on the cable.
So with landable, reusable rockets, the only savings you get from an elevator are :
-
You don’t burn rocket fuel. Which can be liquid methane, which is about $1.52 a gallon.
-
You don’t burn rocket components. Eventually, the reusable rockets will wear out and need to be overhauled or recycled.
I’ve mentioned this in many other threads, but classic science fiction is out-dated. The new thing is that since machine learning allows for making robots that are more flexible and can learn to perform well defined tasks on their own (and learn how to clear faults on their own), manufacturing is about to get a whole lot cheaper. I’m predicting a ‘manufacturing singularity’, actually. Once robotic factories can make the components used in robotic factories, the cost of robotics would plummet to just barely over the cost of raw materials and energy. And you can then have those cheap robots build mining robots, lowering the raw material costs. And you can have the cheap robots manufacture solar panels and batteries for cheap, easy to deploy production of energy.
Anyways, this is immediate term. Better robots are being deployed now and really, really amazing robots are about 4 years away (autonomous cars).
So if we wanted to leave the planet in mass, the way it would be done is that the rocket fuel would be made with solar power to convert air to methane and liquid oxygen. Maybe buy the Sahara desert and cover the whole thing with photovoltaics, and the electricity runs the gas compressors and Sabatier plants when the sun is out, with everything shutting down at night.
Then there would be thousands upon thousands of reusable rockets, all based on common components, all made without a human hand and inspected by machines to a far greater level of precision than current manufacturing tolerances. Most of these rockets would be flying cargo missions every few hours until the end of their service lives (which might be just a few weeks or months before the rockets get recycled and rebuilt in automated factories back to brand new). Over thousands of launches a month, you can get the statistics needed to really verify a launcher’s reliability for crewed payloads.
This is something we can basically start on today. This technology is real. Orbital tethers aren’t possible or feasible.
Both, actually. One fell “more than 15,000 meters” and burnt up upon reentry. Another was flicked off from the counterweight and could not be rescued before her air supply ran out. Here’s a Google Books link to that section of the book.
Liquid methane might be a buck fifty a gallon, but it takes an awful lot of gallons of liquid methane (and of liquid oxygen; don’t forget that part) to lift useful payloads to orbit. And the time might come when we can make methane really cheaply, but if we ever get to that point, we’ll be able to make everything else really cheaply, too, and space travel will still be one of the most expensive things humans do.
And, of course, if you can use your miracle machines to make methane, can you also use them to make nanofiber? Because that really could make space travel cheap, even in comparison to other cheap things.
We’re not talking about ‘miracle machines’. We’re talking about robots doing the same tasks humans are needed to do now. The same robots available now, just optimized for easy of automated assembly and connected to large datacenters to provide their edge case intelligence.
Liquid methane has only a little less energy per gallon than kerosene. Currently, it costs about a million bucks worth of fuel and LOX to put a F9 into orbit. So in this future world of mass production, the cost might be a little lower, and the cost of the rocket itself would be immensely lower, but yeah, it would probably still be hundreds of thousands of dollars per launch.
This type of robots solves none of the problems with space tethers. It doesn’t make it possible to produce the flawless chains of CNT you’d need, nor does it solve the inherent problem of having a cable that is 70,000 kilometers long that if it fails at any point, will negate your entire investment.
This is the reason space tethers won’t work. With thousands of separate rockets made in separate copies of the same automated plant, fueled by distributed arrays of solar panels and gas processing plants, there’s no single point of failure. Any failure only destroys a tiny part of the infrastructure.
Thanks or the confirmation, scr4.
My knowledge of space elevators and orbital mechanics mostly stems from early 80’s hard sci-fi. Glad to know I can still remember the details.
I may need to check out some books from the library. The Descent of Anansi (Barnes/Niven) has nanowires and orbital tether physics, as well.
It’s not about the cost of fuel. As you say, propellent is dirt cheap.
It’s about breaking the delta V budget into smaller chunks and reducing velocity change that needs to be accomplished via reaction mass. The exponent in the rocket equation scales with the delta V budget.
Presently it takes about 9 km/s to get to LEO (Low Earth Orbit) and around 13 km/s for GEO (Geosynchronous Earth Orbit) or TMI (Trans Mars Insertion).
Given a 9 km/s delta V budget and chemical propellent you have a dry mass fraction of around 4%. This 4% includes the payload, structure, rocket engines, power source and propellent tanks.
This tiny mass fraction means upper stages are about as sturdy as a hollowed out egg shell. With this slim mass fraction it hasn’t been doable to give upper stages the structure and thermal protection they need to survive the extreme conditions of an 8 km/s re-entry.
Elon Musk and Jeff Bezos seem well on their way to making reusable boosters. But reusable upper stages are another story.
There’s no reason why a failure at a single point should cause an entire cable to fail. As with anything else humans make, you build in a safety factor. If you make the cable 10 times stronger than it needs to be, then you can survive a cut 90% of the way across the cable. And then you send up a robot to fix the cut, using the same technology you used to make it in the first place.
And of course there’s no reason to only have one space elevator. The marginal cost for making more would be orders of magnitude less than the cost for the first one, because you’re re-using all of the R&D, and you’re using the first one to launch all the rest of them. So even if you do lose one cable entirely, you just use one of the other nine you have to cheaply launch a replacement.
The problem with adding a significant safety factor is the same as that joke about making an airplane out of the same stuff they make black boxes out of. (the highways aren’t wide enough)
It’s likely not physically possible to do that. Adding a safety factor means making the cable thicker, which makes it heavier, which increases the stress, so it doesn’t work.
Sure, if you can build hundreds of orbital tethers, in the far, far future, maybe you’d do it.
Well, then you throw them away. Ultimately, the cost of rockets is tied to the cost of manufacturing. The raw materials and fuel are cheap, as you say.
Does it count as being lost if it is still in your shorts?
That’s what we do. Which is why spaceflight is so expensive. Imagine what a transcontinental flight would cost if we threw away a 747 each trip.
As I said, full blown Clarke Towers aren’t plausible. But their smaller cousins orbital tethers are. And they can lend a hand reducing delta v budgets which could make ships reusable.
Sigh. I am tired of repeating myself. Mass production and automation works. Complex goods are available today for less than the price of much simpler goods in the past. The problem is that mass production requires mass order volumes. What makes it cheaper is that there are a lot of fixed development costs in producing anything, from programming the automation to coming up with the optimal steps for each part of the process to developing the fixtures and custom tooling.
Better automation should lead to robotic systems where you can get the same cost benefits you would get with ordering a million units with ordering only 10. That, in turn, would mean that you in fact could scrap the 747 after each transcontinental flight, and the tickets would only be 10 times as expensive.
Not necessarily. You mass-produce things by making a mold or die, and then using the mold or die to make many parts. The original mold or die can’t be made in the same way, and has to be made in some more expensive way, but that’s OK, because its cost is spread out over many final products.
Correct. But a robot that can flexibly make a lot of things, and any molds needed can just be 3d printed, would be like having a mold or die already made for any order.
Being able to 3D print a mold or die is not the same thing as having a mold or die already made. I’ll be the first to say that 3D printing is cool, but it’s a lot more expensive and time-consuming than mass manufacturing techniques, especially in materials suitable for use as dies or molds.
That might have been true a few years ago. Near instant 3d printers are now available for plastic, and there are extremely high speed metal printers as well. Again, you would print the mold, not the final product.
Maybe if they didn’t show movies and didn’t give out those goldfish crackers?
The “near instant” 3D printers you’re referring to are still several orders of magnitude slower than mass-production methods.