This post is mainly about a design concept for far-future large space habitats.
some proposed designs
As you can see on Wikipedia, many space habitat designs have been proposed. Below are some that I thought were worth mentioning.
current space stations
Obviously, space stations with long-term occupants have already been made, the biggest being the ISS.
issue: small modules
Each launch lifts a complete cylindrical module, and then the modules are assembled. This limits module diameter, which has some problems:
To be able to move things through the center of modules, much of the volume must be left empty.
The large surface area relative to volume makes radiation shielding and thermal insulation less efficient.
Wires and pipes between modules must go through several connections.
ISS maintenance
The ISS has some purposes:
National prestige.
Studying the effects of long-term spaceflight on humans.
Inspiring kids by having humans in space.
International cooperation.
The thing is, (4) is no longer very relevant, and a new space station wouldn't accomplish the other things much better than the existing one. And that's why the ISS is still in orbit past its planned lifetime.
The ISS costs ~$3 billion/year to operate, and it has major maintenance problems. At this point, it might be cheaper to build a new station and abandon the ISS than it is to continue using it, considering that SpaceX is making Starship anyways. NASA seems to agree. Maybe some equipment from the ISS could be brought down and sold to collectors/museums? Some people would probably pay a lot.
Then, if you're making a new station, people might want to see something visibly different. "You want to spend billions of dollars to do the same thing again? We were supposed to learn stuff and make progress from the first time."
inflatable modules
One way to launch modules bigger than the launcher is to make inflatable modules. Multiple companies are working on this, including Axiom Space and Lockheed Martin. This was even tested in space. Here's a video of Bigelow modules.
issue: polymer degradation
Exposed polymers in space degrade fairly quickly. The reason that flexible inflatable modules are maybe practical is, they'd be covered with vacuum insulation using layers of aluminum. Radiation damage would still be a long-term problem, but inflatable modules might last long enough for a space station project.
Spacesuits have used fabric based on glass fiber and teflon largely because it resists damage from the atomic oxygen in space.
issue: bending fabric
The walls of inflatable modules must be thick enough to contain pressure and have insulation. That means the bending radius has to be somewhat large. There's also the possibility of damage where the walls are bent, and the Bigelow module test had some problems inflating because the fabric stuck to itself.
geodesic spheres
Another obvious way to get a space station with more internal volume is to launch a stack of panels that are welded together into a sphere. The panel pattern would be a geodesic polyhedron.
A company called ThinkOrbital has been working on such a design; see this paper. The design in that paper would have 2x the volume of the ISS, from a single launch. It was layer redesigned to use a SpaceX Starship launch and have 2x that volume.
ThinkOrbital's approach involves metal panels welded together using a robotic arm with an electron beam welder. The basic concept seems entirely feasible to me. Many people guess that doing welding in space is a big problem, but it was done by the Soviet Union. I personally like this concept better than inflatable modules.
comments on ThinkOrbital
ThinkOrbital proposed using aluminum alloys, but aluminum welds are weak. So, they proposed panels that are thicker at welds, but another solution is to use steel or titanium instead, at least at the panel edges. Titanium should also warp less from welding than aluminum. To me, titanium seems like a better option for panels here than aluminum.
ThinkOrbital specifies panels as big as the launcher can hold, to minimize welding. With titanium panels, maybe the panels should be smaller to make the handling equipment lighter and reduce tooling costs.
issue: insulation gaps
If the insulation is integrated into the panels, then there will be gaps in the insulation where panels meet. Obviously there are ways to deal with this, but that adds a little complexity.
issue: no gravity
Of course, current space stations have no gravity. This is useful in some ways, but also causes a lot of problems: health problems, particles floating around, lack of convection, etc.
rotating wheels
So, you want gravity, which means rotation, and you probably want to be able to roll stuff around the station. The obvious solution to that is a rotating ring, which was proposed back in 1903 by Tsiolkovsky. Here's Wikipedia on that, and here's an old example design.
issue: radiation shielding
Having a lot of surface area relative to volume makes radiation shielding harder. How much of a problem is radiation in space?
Forty-eight cases of severe lens opacification (16.2%) were observed among the 295 NASA astronauts who participated in the LSAH (Longitudinal Study of Astronaut Health) study, but 86% of the astronauts who stayed in space suffered from a pathology of the eye.
The bone loss seen in astronauts has generally been attributed just to lack of gravity, but I think radiation is responsible for a significant fraction of it.
That's from being in low earth orbit, and even in LEO the Earth's magnetic field provides significant protection. In interstellar space, at the same distance from the sun, radiation doses are ~1000x higher. Radiation shielding must then be at least 10mm thick even for short-term usage.
issue: large-scale industry
Nobody's building large space habitats anytime soon, but here's we're considering what would be necessary for a self-sustaining space-based civilization, and the economy for that would probably require a large scale. Industry often has large minimum scales for manufacturing things efficiently. If production is distributed across many small wheels, transporting items between them is difficult.
big rings
If radiation shielding and industry minimum scales are a problem, let's try going bigger. What if we increase the ring diameter?
The ultimate example of that would be a Ringworld: a ring all the way around a sun. That's been seen in some SF but it's rather impractical.
issue: orbit stability
The orbit of a ring around a sun isn't stable. Any perturbation will increase over time until the ring breaks apart.
So, let's consider a much smaller ring that orbits around a sun, instead of going all the way around a sun, but is still large enough for a complete economy.
issue: material strength
The Ringworld book has the ring rotating faster than its orbital speed to provide gravity. The thing is, for a given amount of centrifugal force, required material strength is proportional to radius. You really want a diameter less than a few kilometers for 1g of gravity.
issue: lack of modularity
How do you gradually construct a very large ring while using it? That seems difficult.
issue: material transport
Transporting materials inside a large space station with no gravity is easy. With no weight to support, payloads can simply float towards their destination.
With a large-diameter ring, materials must be moved inward a long distance to make them weightless. Axial transport is then easy, but there's only a short axial distance to travel. Obviously, it's possible to transport materials on wheeled vehicles, but we'd prefer not to need to.
long cylinders
OK, large-diameter rings require strong materials and are hard to make. So, let's try increasing length instead.
Arthur C. Clarke wrote about a large rotating cylindrical space habitat back in 1973.
issue: rotation stability
When a narrow object is spun along its long axis, its rotation is not stable. If there's some energy dissipation from flexing or sloshing, its rotation eventually changes to its axis of greatest rotational inertia; that's the Dzhanibekov effect. So, a cylinder would end up spinning the wrong way.
Consider holding a piece of cooked spaghetti from one end and spinning it along its long axis; the spaghetti won't stay straight. Very long and thin cylinders could have a similar problem.
issue: bearings
If a pair of counter-rotating cylinders is used, they need to be connected by bearings.
It's not practical to perfectly balance those cylinders, so they'll tend to wobble slightly. So, bearings between them must be flexible enough to handle some relative movement. If the wobbling tendencies of rotating cylinders are cancelled out with counter-rotation, then the bearings must exert enough force to do that, and they'd probably need active control. The longer the cylinders, the more torque must be exerted on them to keep them balanced.
big rotating spheres
If a big ring is too wide and a rotating cylinder isn't stable, we could compromise and use a big rotating sphere. That's a "Bernal sphere". Obviously, that can't be as big as a very long cylinder, but let's suppose industry could be streamlined to the point where a whole economy could fit in one of those. (If you can accept 0.5g gravity, then the diameter could be doubled.)
If you're looking for a much bigger sphere for a SF story, maybe what you want is a bubbleworld.
issue: no microgravity zone
There's no large volume where materials can be floated around freely. Having that is useful for industry and transportation.
issue: no doors
Most proposed large cylindrical space habitats have large open areas. If something causes a large air leak in one section, it would then be impossible to contain the leak. Of course, doors could be added, but they'd need to contain the air pressure. So, section doors would need to be large hemispheres, which would take up a lot of space. That's at least possible for cylinders, but with a sphere, there's really no way to add internal bulkheads.
issue: lack of modularity
A long cylinder can be gradually extended, but expanding a sphere gradually is much harder.
bhabitats
Above, we ruled out small disconnected habitats, large-diameter rotating rings, and very long rotating cylinders. So, to get some gravity, the only remaining option is many small rotating structures connected together in a way that allows for easy material transport.
Here's a space habitat design I made on that basis. For now, I'm calling this type "bhabitats". The basic concept is: many pressurized spheres that each connect to other spheres and to rotating cylinders, in a modular way.
diagram
Here's a diagram of a large assembled bhabitat:
expansion steps
first small ball
Make a sphere of maybe 600m diameter, maybe out of iron. It might be made by welding polygonal panels together.
The sphere should have 8 locations where ports can be added: 6 small ports and 2 medium ports. Small ports allow for relative rotation.
Fill the sphere with air. Now, there's a pressurized environment with a little bit of radiation shielding.
make some shielding
There are 2 things you want for radiation shielding: a magnetic field, and lots of mass. Here, a magnetic field means superconducting coils. Ideally, most of the shielding mass would be easily-available material, such as rocks collected in space.
Make a big superconducting coil that goes around the habitat. Also make some movable panels containing rocks, and place them around the habitat.
Now, there's a pressurized environment with better radiation shielding.
first cylinder
Make a cylinder of maybe 400m diameter with hemispherical ends, maybe 800m long in total. Connect one cylinder end to a small port of the small ball. Spin the cylinder and small ball in opposite directions.
Now, there's some living space with gravity.
2 cylinders
Connect a 2nd cylinder to the small ball on the opposite side of the 1st, and rotating in the opposite direction.
Now, there's no need for the small ball to rotate. So, shielding panels can be attached to the small ball.
3 cylinders
Connect a 3rd cylinder, for 3 in total. Now, the habitat can freely change its orientation by adjusting rotation rates of cylinders.
first cylinder star
Connect 3 more cylinders to the small ball, for 6 in total.
first cylinder string
Connect the small balls of multiple cylinder stars, using their medium ports.
first big ball
At the end of the cylinder string, make a sphere of maybe 2km diameter. It should have 1 medium port to the cylinder string, and locations for 2 large ports for connections to other big balls.
Now, the habitat has a large microgravity area for industry, which also provides shielding from that direction.
main string
Connect big balls together using the large ports. Optionally, connect the cylinder strings with a truss for structural strength.
To make space for the cylinder strings, the big balls need to be rotated somewhat unless they're very large. The above diagram shows a partial ring of cylinder strings; another option is cylinder strings on alternating sides, but that would've made for a bigger diagram.
Yes, it's possible to make a 2d grid of big balls, but that wouldn't give enough area for solar panels and radiators.
component design
cylinders
The above sections described 400M diameter cylinders. Such cylinders might have 20 floors, ranging from 1g to 0.5g, with a hollow section in the center.
At 400M diameter, the material requirements for 1g are very reasonable and the structure can be lightweight. One reason for larger diameters being proposed is concerns about motion sickness, but I think that, like with boats and VR movement, people would get used to it. Certainly, you could go bigger without making structural mass a big problem, but a smaller diameter has some advantages for internal transportation. Shorter elevators to the center are better, and very thick cylinders could have elevator capacity problems, much like skyscrapers do today. Another issue with very large cylinders is, they'd need larger linear actuators at the port for a given amount of wobbling.
To maintain mass balance, the cylinders need some movable masses. The balancing system might use a:
2 rings of rails around the cylinder, 1 near each end.
A set of heavy vehicles that travel around the rail rings to balance the cylinder.
small ports
The bearings can use an ionic liquid layer or gallium alloy to contain the air.
Some cylinder wobbling is inevitable. You need:
flexible connections to the bearing, probably corrugated tubes
active damping with linear actuators
electrical connections
Electric power can be transferred through liquid metal sliding contacts, maybe using both NaK eutectic and gallium alloy. Those are used today for applications including wind turbines and rotating radars.
We can send electricity through an arm to the far side of the cylinder, and have the rotating electric contacts near the center.
internal transportation
Most transportation would start with an elevator towards the center of a cylinder. At the center, cargo could be moved with propellers and small wings, like airships without the gas bladders. Near the center, thrower/catcher devices that can handle 15 mph could launch payloads anywhere along the cylinder, without occupying the center path used for traffic to/from the port.
This post is mainly about a design concept for far-future large space habitats.
some proposed designs
As you can see on Wikipedia, many space habitat designs have been proposed. Below are some that I thought were worth mentioning.
current space stations
Obviously, space stations with long-term occupants have already been made, the biggest being the ISS.
issue: small modules
Each launch lifts a complete cylindrical module, and then the modules are assembled. This limits module diameter, which has some problems:
To be able to move things through the center of modules, much of the volume must be left empty.
The large surface area relative to volume makes radiation shielding and thermal insulation less efficient.
Wires and pipes between modules must go through several connections.
ISS maintenance
The ISS has some purposes:
The thing is, (4) is no longer very relevant, and a new space station wouldn't accomplish the other things much better than the existing one. And that's why the ISS is still in orbit past its planned lifetime.
The ISS costs ~$3 billion/year to operate, and it has major maintenance problems. At this point, it might be cheaper to build a new station and abandon the ISS than it is to continue using it, considering that SpaceX is making Starship anyways. NASA seems to agree. Maybe some equipment from the ISS could be brought down and sold to collectors/museums? Some people would probably pay a lot.
Then, if you're making a new station, people might want to see something visibly different. "You want to spend billions of dollars to do the same thing again? We were supposed to learn stuff and make progress from the first time."
inflatable modules
One way to launch modules bigger than the launcher is to make inflatable modules. Multiple companies are working on this, including Axiom Space and Lockheed Martin. This was even tested in space. Here's a video of Bigelow modules.
issue: polymer degradation
Exposed polymers in space degrade fairly quickly. The reason that flexible inflatable modules are maybe practical is, they'd be covered with vacuum insulation using layers of aluminum. Radiation damage would still be a long-term problem, but inflatable modules might last long enough for a space station project.
Spacesuits have used fabric based on glass fiber and teflon largely because it resists damage from the atomic oxygen in space.
issue: bending fabric
The walls of inflatable modules must be thick enough to contain pressure and have insulation. That means the bending radius has to be somewhat large. There's also the possibility of damage where the walls are bent, and the Bigelow module test had some problems inflating because the fabric stuck to itself.
geodesic spheres
Another obvious way to get a space station with more internal volume is to launch a stack of panels that are welded together into a sphere. The panel pattern would be a geodesic polyhedron.
A company called ThinkOrbital has been working on such a design; see this paper. The design in that paper would have 2x the volume of the ISS, from a single launch. It was layer redesigned to use a SpaceX Starship launch and have 2x that volume.
ThinkOrbital's approach involves metal panels welded together using a robotic arm with an electron beam welder. The basic concept seems entirely feasible to me. Many people guess that doing welding in space is a big problem, but it was done by the Soviet Union. I personally like this concept better than inflatable modules.
comments on ThinkOrbital
ThinkOrbital proposed using aluminum alloys, but aluminum welds are weak. So, they proposed panels that are thicker at welds, but another solution is to use steel or titanium instead, at least at the panel edges. Titanium should also warp less from welding than aluminum. To me, titanium seems like a better option for panels here than aluminum.
ThinkOrbital specifies panels as big as the launcher can hold, to minimize welding. With titanium panels, maybe the panels should be smaller to make the handling equipment lighter and reduce tooling costs.
issue: insulation gaps
If the insulation is integrated into the panels, then there will be gaps in the insulation where panels meet. Obviously there are ways to deal with this, but that adds a little complexity.
issue: no gravity
Of course, current space stations have no gravity. This is useful in some ways, but also causes a lot of problems: health problems, particles floating around, lack of convection, etc.
rotating wheels
So, you want gravity, which means rotation, and you probably want to be able to roll stuff around the station. The obvious solution to that is a rotating ring, which was proposed back in 1903 by Tsiolkovsky. Here's Wikipedia on that, and here's an old example design.
issue: radiation shielding
Having a lot of surface area relative to volume makes radiation shielding harder. How much of a problem is radiation in space?
This paper notes:
The bone loss seen in astronauts has generally been attributed just to lack of gravity, but I think radiation is responsible for a significant fraction of it.
That's from being in low earth orbit, and even in LEO the Earth's magnetic field provides significant protection. In interstellar space, at the same distance from the sun, radiation doses are ~1000x higher. Radiation shielding must then be at least 10mm thick even for short-term usage.
issue: large-scale industry
Nobody's building large space habitats anytime soon, but here's we're considering what would be necessary for a self-sustaining space-based civilization, and the economy for that would probably require a large scale. Industry often has large minimum scales for manufacturing things efficiently. If production is distributed across many small wheels, transporting items between them is difficult.
big rings
If radiation shielding and industry minimum scales are a problem, let's try going bigger. What if we increase the ring diameter?
The ultimate example of that would be a Ringworld: a ring all the way around a sun. That's been seen in some SF but it's rather impractical.
issue: orbit stability
The orbit of a ring around a sun isn't stable. Any perturbation will increase over time until the ring breaks apart.
So, let's consider a much smaller ring that orbits around a sun, instead of going all the way around a sun, but is still large enough for a complete economy.
issue: material strength
The Ringworld book has the ring rotating faster than its orbital speed to provide gravity. The thing is, for a given amount of centrifugal force, required material strength is proportional to radius. You really want a diameter less than a few kilometers for 1g of gravity.
issue: lack of modularity
How do you gradually construct a very large ring while using it? That seems difficult.
issue: material transport
Transporting materials inside a large space station with no gravity is easy. With no weight to support, payloads can simply float towards their destination.
With a large-diameter ring, materials must be moved inward a long distance to make them weightless. Axial transport is then easy, but there's only a short axial distance to travel. Obviously, it's possible to transport materials on wheeled vehicles, but we'd prefer not to need to.
long cylinders
OK, large-diameter rings require strong materials and are hard to make. So, let's try increasing length instead.
Arthur C. Clarke wrote about a large rotating cylindrical space habitat back in 1973.
issue: rotation stability
When a narrow object is spun along its long axis, its rotation is not stable. If there's some energy dissipation from flexing or sloshing, its rotation eventually changes to its axis of greatest rotational inertia; that's the Dzhanibekov effect. So, a cylinder would end up spinning the wrong way.
O'Neill proposed a pair of counter-rotating cylinders. That mitigates the rotation stability issue, but then...
issue: rigidity
Consider holding a piece of cooked spaghetti from one end and spinning it along its long axis; the spaghetti won't stay straight. Very long and thin cylinders could have a similar problem.
issue: bearings
If a pair of counter-rotating cylinders is used, they need to be connected by bearings.
It's not practical to perfectly balance those cylinders, so they'll tend to wobble slightly. So, bearings between them must be flexible enough to handle some relative movement. If the wobbling tendencies of rotating cylinders are cancelled out with counter-rotation, then the bearings must exert enough force to do that, and they'd probably need active control. The longer the cylinders, the more torque must be exerted on them to keep them balanced.
big rotating spheres
If a big ring is too wide and a rotating cylinder isn't stable, we could compromise and use a big rotating sphere. That's a "Bernal sphere". Obviously, that can't be as big as a very long cylinder, but let's suppose industry could be streamlined to the point where a whole economy could fit in one of those. (If you can accept 0.5g gravity, then the diameter could be doubled.)
If you're looking for a much bigger sphere for a SF story, maybe what you want is a bubbleworld.
issue: no microgravity zone
There's no large volume where materials can be floated around freely. Having that is useful for industry and transportation.
issue: no doors
Most proposed large cylindrical space habitats have large open areas. If something causes a large air leak in one section, it would then be impossible to contain the leak. Of course, doors could be added, but they'd need to contain the air pressure. So, section doors would need to be large hemispheres, which would take up a lot of space. That's at least possible for cylinders, but with a sphere, there's really no way to add internal bulkheads.
issue: lack of modularity
A long cylinder can be gradually extended, but expanding a sphere gradually is much harder.
bhabitats
Above, we ruled out small disconnected habitats, large-diameter rotating rings, and very long rotating cylinders. So, to get some gravity, the only remaining option is many small rotating structures connected together in a way that allows for easy material transport.
Here's a space habitat design I made on that basis. For now, I'm calling this type "bhabitats". The basic concept is: many pressurized spheres that each connect to other spheres and to rotating cylinders, in a modular way.
diagram
Here's a diagram of a large assembled bhabitat:
expansion steps
first small ball
Make a sphere of maybe 600m diameter, maybe out of iron. It might be made by welding polygonal panels together.
The sphere should have 8 locations where ports can be added: 6 small ports and 2 medium ports. Small ports allow for relative rotation.
Fill the sphere with air. Now, there's a pressurized environment with a little bit of radiation shielding.
make some shielding
There are 2 things you want for radiation shielding: a magnetic field, and lots of mass. Here, a magnetic field means superconducting coils. Ideally, most of the shielding mass would be easily-available material, such as rocks collected in space.
Make a big superconducting coil that goes around the habitat. Also make some movable panels containing rocks, and place them around the habitat.
Now, there's a pressurized environment with better radiation shielding.
first cylinder
Make a cylinder of maybe 400m diameter with hemispherical ends, maybe 800m long in total. Connect one cylinder end to a small port of the small ball. Spin the cylinder and small ball in opposite directions.
Now, there's some living space with gravity.
2 cylinders
Connect a 2nd cylinder to the small ball on the opposite side of the 1st, and rotating in the opposite direction.
Now, there's no need for the small ball to rotate. So, shielding panels can be attached to the small ball.
3 cylinders
Connect a 3rd cylinder, for 3 in total. Now, the habitat can freely change its orientation by adjusting rotation rates of cylinders.
first cylinder star
Connect 3 more cylinders to the small ball, for 6 in total.
first cylinder string
Connect the small balls of multiple cylinder stars, using their medium ports.
first big ball
At the end of the cylinder string, make a sphere of maybe 2km diameter. It should have 1 medium port to the cylinder string, and locations for 2 large ports for connections to other big balls.
Now, the habitat has a large microgravity area for industry, which also provides shielding from that direction.
main string
Connect big balls together using the large ports. Optionally, connect the cylinder strings with a truss for structural strength.
To make space for the cylinder strings, the big balls need to be rotated somewhat unless they're very large. The above diagram shows a partial ring of cylinder strings; another option is cylinder strings on alternating sides, but that would've made for a bigger diagram.
Yes, it's possible to make a 2d grid of big balls, but that wouldn't give enough area for solar panels and radiators.
component design
cylinders
The above sections described 400M diameter cylinders. Such cylinders might have 20 floors, ranging from 1g to 0.5g, with a hollow section in the center.
At 400M diameter, the material requirements for 1g are very reasonable and the structure can be lightweight. One reason for larger diameters being proposed is concerns about motion sickness, but I think that, like with boats and VR movement, people would get used to it. Certainly, you could go bigger without making structural mass a big problem, but a smaller diameter has some advantages for internal transportation. Shorter elevators to the center are better, and very thick cylinders could have elevator capacity problems, much like skyscrapers do today. Another issue with very large cylinders is, they'd need larger linear actuators at the port for a given amount of wobbling.
To maintain mass balance, the cylinders need some movable masses. The balancing system might use a:
small ports
The bearings can use an ionic liquid layer or gallium alloy to contain the air.
Some cylinder wobbling is inevitable. You need:
electrical connections
Electric power can be transferred through liquid metal sliding contacts, maybe using both NaK eutectic and gallium alloy. Those are used today for applications including wind turbines and rotating radars.
We can send electricity through an arm to the far side of the cylinder, and have the rotating electric contacts near the center.
internal transportation
Most transportation would start with an elevator towards the center of a cylinder. At the center, cargo could be moved with propellers and small wings, like airships without the gas bladders. Near the center, thrower/catcher devices that can handle 15 mph could launch payloads anywhere along the cylinder, without occupying the center path used for traffic to/from the port.