How the heck do you have gas-tight bearings between non rotating and rotating sections?

[bonyetty]

Seven small ones just on the ISS cupola. 

https://m.youtube.com/watch?v=KSzuiqVjJg4

[SomeGuy123]

3 hours ago, Nibb31 said:

The Apollo stack used "barbecue roll" spin on its way to the Moon, not for gravity but for thermal regulation. The fuel consumption to spin up Apollo was negligeable compared to the size of the spacecraft. I don't see why it wouldn't scale gracefully to a larger vehicle. You could even use SEP thrusters to do it on a larger craft. And the induced spin stabilization actually saves propellant used for attitude corrections.

You know what would make everything heavier? Joints, connections, seals, motors, and everything needed for a spin section. It would also make everything more complex and more failure prone. What happens if it jams? What happens if it leaks? What happens if you run out of lubricant?

What you're neglecting is that when you spin something fast enough that there's enough apparent gravity/centripetal acceleration at the end to stand up and walk around (so at least 1/6 G, probably more), the structure of the spacecraft in that section has to withstand stress equal to the 'weight' of the spun sections.

That structure is mass you wouldn't otherwise need.  I'm describing interplanetary vehicles that are made in orbit and will never leave a low gravity environment.  They will never touch a planet's atmosphere, and their engines won't subject their structures to more than 1/100 of 1 gravity (10 centimeters/second squared) even with nearly dry fuel tanks.  

Things like the heat radiators are optimal if they are incredibly thin foil sheets or vapor sprays.  Fuel tanks are optimal if they are basically balloons just strong enough to keep the slush hydrogen inside in one piece.  And so on.

By making everything strong enough to be spun you're making everything heavier, increasing dry mass, and thus reducing your performance and increasing the minimum propellant needed for a given mission.  

As for complexity : it's relative.  A nuclear interplanetary spacecraft won't be simple, rotating bearings or not, and the bearings are not the complex part, either.  A lightweight heat radiator system or a fusion engine would be immensely more complex.  By making it lighter, you reduce costs, leaving your more money for your immensely complex engines and heat rejection systems and closed loop life support systems and so on.

Edited February 2, 2016 by SomeGuy123

[SomeGuy123]

6 hours ago, Elthy said:

Another problem with only parts of the craft rotating: How do you transfer energy, fluids, gasses between the parts? Energy is propably easy, but the others...

I'd want to do it with omniporters.

These are robots that drive along grooves or rails in the outer hull of the ship, or transit down conveyer tubes.  They carry roughly a cube shaped "cargo pod" that can have anything in it that will fit.

Well, there would be cryogenic gas pods - a cryogenic bottle shaped to fit inside a cube, with various ports that robots can connect to.  So to transfer between spinning and non spinning sections, one of the main things to move is cryogenic gas.  So the liquid CO2/liquid O2 would be used to fill the gas pods, then porters would move along the grooves to the interface between spinning and non spin sections.  There would be an outer transfer ring with a groove in it.  Porter moves into that groove.

From non-spin section -> spin : transfer ring matches speed with the spin section (the transfer ring is mounted in the non spin section), so the groove is now perfectly matched with the groove in the spin section.

From spin -> nonspin.  Robot waits in the last groove segment of the spin section until the transfer ring matches speed.  Then it slides over into the groove on the transfer ring.  Transfer ring decelerates to rest and the robot proceeds on it's way.

This to me makes sense.  You might bring aboard asteroid ore, each section of ore kept in a separate cube storage container in a holding bay.  You send the ore to the smelter and the output ingots get carried away in one set of porters while the slag gets stored in a separate set of cube containers.

So the ship's raw materials are kept in containers that are separate now.  When a manufacturing order is placed, a whole lineup of porters bring by cubes containing each needed material.  More realistically you'd probably end up storing many intermediate stages as you build things.  

 

[KerikBalm]

I can imagine a number of possibilities for long term space travel.

#1) Has already been suggested, is that only the interior rotates within the stationary exterior donut. Just basically have the interrior "donut" on a maglev track, and its just air friction rather than bearing friction... very low power consumption

#2) Similar to rotating the entire craft... you just don't have an airtight seal. The rotating section has is sealed, and there is no airtight connection to the non rotating section. If you need to enter the non rotating section, you have to stop all rotation, and connect the airlocks. In this case, to save reaction mass if you do that alot, then youd have counter rotating sections, each sealed internally. You could spin them up or down simultaneously for 0 net torque. In this case, you'd be unable to access the other parts of the ship except when the rotating sections are "spun down" and linked by a non-rotating airlock.

You could still move between 1g and zero g at will by simply moving radially towards the center.

[Nibb31]

 

Quote

Things like the heat radiators are optimal if they are incredibly thin foil sheets or vapor sprays.  Fuel tanks are optimal if they are basically balloons just strong enough to keep the slush hydrogen inside in one piece.  And so on.

I was considering near term technology this thread was about near-term technology that might be needed in 20 or 30 years. If we are going into Star Trek grade fantasy with orbital dry docks and vapor spray radiator material, then pretty much anything is possible, including 1G brachistochrone gravity. In that universe, a few tons of extra weight is the least of your troubles.

For any interplanetary mission that we are likely to see in the foreseeable future, we are stuck with orbital assembly of modules that are built on the ground. I'm thinking more about The Martian's Hermes, 2001's Discovery or Nautilus-X technology than USS Enterprise or Battlestar Galactica. Anything built in a 1G environment and designed to be launched on a rocket is likely to be able to cope with a reasonable amount of artificial gravity in space without much extra weight.

If we ever get to the point of star docks and asteroid mining, then technology will have advanced to a point where the discussion we are having today will be moot, so there isn't much point going there.

Quote

By making everything strong enough to be spun you're making everything heavier, increasing dry mass, and thus reducing your performance and increasing the minimum propellant needed for a given mission.  

As for complexity : it's relative.  A nuclear interplanetary spacecraft won't be simple, rotating bearings or not, and the bearings are not the complex part, either.  A lightweight heat radiator system or a fusion engine would be immensely more complex.  By making it lighter, you reduce costs, leaving your more money for your immensely complex engines and heat rejection systems and closed loop life support systems and so on.

Even if those systems are complex, you still want to avoid as many moving parts as possible. Anything that moves is going to produce friction, wear, and heat. Therefore it's simply more likely to cause problems.

Edited February 2, 2016 by Nibb31

[fredinno]

1 hour ago, Nibb31 said:

Any extra structure is easily offset by the extra 

I think that any joints or 

I was considering near term technology this thread was about near-term technology that might be needed in 20 or 30 years. If we are going into Star Trek grade fantasy with orbital dry docks and vapor spray radiator material, then pretty much anything is possible, including 1G brachistochrone gravity. In that universe, a few tons of extra weight is the least of your troubles.

For any interplanetary mission that we are likely to see in the foreseeable future, we are stuck with orbital assembly of modules that are built on the ground. I'm thinking more about The Martian's Hermes, 2001's Discovery or Nautilus-X technology than USS Enterprise or Battlestar Galactica. Anything built in a 1G environment and designed to be launched on a rocket is likely to be able to cope with a reasonable amount of artificial gravity in space without much extra weight.

If we ever get to the point of star docks and asteroid mining, then technology will have advanced to a point where the discussion we are having today will be moot, so there isn't much point going there.

Even if those systems are complex, you still want to avoid as many moving parts as possible. Anything that moves is going to produce friction, wear, and heat. Therefore it's simply more likely to cause problems.

Nibb, he WAS talkgin about near-term. Balloon tanks are current-technology level since Atlas was made in the sixties.

[Nibb31]

6 minutes ago, fredinno said:

Nibb, he WAS talkgin about near-term. Balloon tanks are current-technology level since Atlas was made in the sixties.

I know, and they seem to handle 1G gravity (and several Gs of acceleration) pretty well AFAIK.

[DerekL1963]

11 hours ago, Nibb31 said:

Slush isn't much of a problem once it's rotating. Just add some baffles in the tank. If anything, gravity induced ullage is a benefit. 

No, it's not much of a problem once you're rotating.   The problems come during spin-up and spin-down.

And how exactly is gravity induced ullage a benefit?   Spin ullage pastes the fuel to the walls of the tanks, and when you spin down to maneuver it won't be pasted to the walls anymore.

11 hours ago, Nibb31 said:

The fuel consumption to spin up Apollo was negligeable compared to the size of the spacecraft.

The Apollo spacecraft spun very, very slooooooowwwwly. Around .1RPH(our).   A very small fraction of the speed required to generate even fractional-G (say lunar gravity) in any reasonable sized spacecraft.

 

[Nibb31]

 

22 minutes ago, DerekL1963 said:

No, it's not much of a problem once you're rotating.   The problems come during spin-up and spin-down.

Still not a huge deal. We have know how to make tank baffles for decades. Otherwise, airplanes would be falling out of the sky from fuel slosh each time they turn.

Quote

And how exactly is gravity induced ullage a benefit?   Spin ullage pastes the fuel to the walls of the tanks, and when you spin down to maneuver it won't be pasted to the walls anymore.

Any ullage is better than no ullage. If you know that the propellant is going to be sticking to the walls of the tanks, that's where you put your fuel outlet.

And who says you have to spin down for maneuvers ?

Edited February 2, 2016 by Nibb31

[DerekL1963]

3 hours ago, Nibb31 said:

Still not a huge deal. We have know how to make tank baffles for decades. Otherwise, airplanes would be falling out of the sky from fuel slosh each time they turn.

0.o   If the problems were even remotely comparable, you'd have a point.  (Hint:  Airplanes don't start with fuel floating freely in their tanks.)

 

[Nibb31]

Jet fighters pull negative and positive Gs, at some point in between, they are at zero-G. The same is true for Vomit Comet aircraft. Slosh is slosh. It's far from an unsolvable problem from an engineering standpoint with some clever use of the baffles, bladders or membranes.

I really don't see what the big deal is. If you can't design something as simple as a tank with an anti-slosh system, then there is no point in even trying to go to Mars in a spacecraft with a rotating section and pressurized joints.

[shynung]

Don't forget bladder tanks! 

[KSK]

Hang on.  If you spin up a habitation ring isn't the rest of the spacecraft is going to spin up in the opposite direction anyway? It might not end up spinning as fast depending on the mass ratio between the ring and the rest of the spacecraft but you'd still need to design the whole thing to cope with spin.

I suppose you could counter that with thrusters during spin-up, which would reduce the slosh problem but you'd still be applying a torque across the spacecraft and it would still need to be sturdy enough to handle that.

Or am I missing something?

[SomeGuy123]

5 hours ago, KSK said:

Hang on.  If you spin up a habitation ring isn't the rest of the spacecraft is going to spin up in the opposite direction anyway? It might not end up spinning as fast depending on the mass ratio between the ring and the rest of the spacecraft but you'd still need to design the whole thing to cope with spin.

I suppose you could counter that with thrusters during spin-up, which would reduce the slosh problem but you'd still be applying a torque across the spacecraft and it would still need to be sturdy enough to handle that.

Or am I missing something?

You aren't, but there are better solutions that don't consume propellant.  The optimizer in me says the most mass-efficient way to handle this is to divide the habitat ring into 2 separate sections.  Try to make the split such that there isn't the need for an excessive amount of traffic between the 2 rings.  For example, processing fluids in zero g is annoying, and some terrestrial plants don't like to grow, so maybe the food production and water recycling systems are in ring B while in ring A you have the crew.  Or maybe the split is that ring A is sleeping areas and residential and ring B is actual work areas.  

Anyways, however you do it, the rings spin in opposite directions and have the same mass.  So the net angular momentum adds to zero.

One way to switch between rings that eliminates most or all air loss is to have a transfer ring.  This is a ring that can spin up or spin down, and it is located between the 2 main rings.  It might actually not be a complete ring, just 2 dumbbells attached to a central hub.  To transfer it speeds up to match velocity with one ring and connects hatches to the side.  Crew and cargo get in.  It then spins down and spins up the other direction.  (a lot of acceleration if you do this quick, so there might be acceleration couches)

A transfer ring design avoids hub bearings completely, the hub would just be a central point where all the tension cables supporting the weight of the ring come together (and an electromagnetic bearing).

Even with a transfer ring, you'd need a fourth, "counterweight" ring.  This ring would be full of weights.  You could actually make it a solid disk made of tungsten or lead, and have it dual purpose, acting like a radiation shield between the crew section and the engine.  

This right might rotate clockwise or counterclockwise, at a wide variety of velocities, including some that would crush humans from g forces if they were on the ring.  

This ring would act to balance the angular momentum between the 3 rings (main ring A, main ring B, transfer ring) so the net actually adds to zero.  That way the 2 main rings can remain at constant velocity.

You could also use this ring as a gigantic gyroscope to force the ship to spin on an axis.

 

 

[KSK]

OK but I think we're definitely starting to get away from any advantages that a habitation ring provides, versus just spinning the ship as advocated by Nibb31. It strikes me that a ship would need to be pretty sturdy to support multiple ring hubs, let alone that counterweight ring. At which point you may as well make the whole ship spinnable and do away with a lot of complexity.

[SomeGuy123]

12 minutes ago, KSK said:

OK but I think we're definitely starting to get away from any advantages that a habitation ring provides, versus just spinning the ship as advocated by Nibb31. It strikes me that a ship would need to be pretty sturdy to support multiple ring hubs, let alone that counterweight ring. At which point you may as well make the whole ship spinnable and do away with a lot of complexity.

No.  There's still an immense difference in complexity between "3 or 4 ring modules with the same thing in each" and "everything in the ship must be designed around always being spun."

The facts of the problem tell me that Nibb31 is flat out wrong, his concept is roughly as bad as the Apollo proposals before they settled on separating the lander and command modules.  

Edited February 3, 2016 by SomeGuy123

[KSK]

Sorry, I disagree but it might help if I set out how I envisage this thing looking. Basically, I'm picturing a relatively thin central spine with the engines on one end and hubs for your habitation rings in the middle. The habitation rings are mounted on the hubs - they could be discs, they could be spoked wheels - doesn't matter much for this discussion. Fuel tanks and other ship systems are strung along the spine.

Now, the spine is the load bearing bit, so your fuel tanks can be fairly light. As Nibb pointed out, anything capable of getting out a gravity well should be just fine. Fuel slosh might be a problem but as already discussed, there are ways around that problem. The spine needs to be fairly light too but it also needs to be stiff enough to support one or more habitation rings. KSP style wet noodle constructions are right out, especially if your habitation rings are spinning because you'll not want those big spinning rings to be wobbling about.

I really don't see why that spine + engines + tanks construction would have a major problem with being spun about it's long axis. If it's stiff enough to support a pair of spinning wheels, it's stiff enough to support being spun itself. The habitation rings are a different matter - they need to be light to minimise the ship's moment of inertia so that you can spin them, or the ship, as easily as possible. On the other hand they also need to be strong enough to withstand their own rotation. But you need to overcome that problem whether you're spinning the whole ship or just spinning the hab rings.

Given all of the above, I think that having the entire ship spinning is the easiest, simplest and safest option and quite possibly the lightest given that you don't need the added mass of hab ring bearings or, using your example, counterweight rings. I don't see what advantage having the rings spinning separately from the ship provides and I see a whole lot of problems if anything goes wrong with one of your bearings.

So yeah - I'm still with Nibb31 on this one.

Edited February 3, 2016 by KSK

[fredinno]

3 hours ago, KSK said:

Sorry, I disagree but it might help if I set out how I envisage this thing looking. Basically, I'm picturing a relatively thin central spine with the engines on one end and hubs for your habitation rings in the middle. The habitation rings are mounted on the hubs - they could be discs, they could be spoked wheels - doesn't matter much for this discussion. Fuel tanks and other ship systems are strung along the spine.

Now, the spine is the load bearing bit, so your fuel tanks can be fairly light. As Nibb pointed out, anything capable of getting out a gravity well should be just fine. Fuel slosh might be a problem but as already discussed, there are ways around that problem. The spine needs to be fairly light too but it also needs to be stiff enough to support one or more habitation rings. KSP style wet noodle constructions are right out, especially if your habitation rings are spinning because you'll not want those big spinning rings to be wobbling about.

I really don't see why that spine + engines + tanks construction would have a major problem with being spun about it's long axis. If it's stiff enough to support a pair of spinning wheels, it's stiff enough to support being spun itself. The habitation rings are a different matter - they need to be light to minimise the ship's moment of inertia so that you can spin them, or the ship, as easily as possible. On the other hand they also need to be strong enough to withstand their own rotation. But you need to overcome that problem whether you're spinning the whole ship or just spinning the hab rings.

Given all of the above, I think that having the entire ship spinning is the easiest, simplest and safest option and quite possibly the lightest given that you don't need the added mass of hab ring bearings or, using your example, counterweight rings. I don't see what advantage having the rings spinning separately from the ship provides and I see a whole lot of problems if anything goes wrong with one of your bearings.

So yeah - I'm still with Nibb31 on this one.

Spinning the rings also gives you a energy storing option via flywheels. Also, you save propellant needed for constantly de-spinning every time you want to do a small burn.

[sgt_flyer]

one simple possibility, if you want to have both rotating and non rotating sections on a spacecraft, (notably to easily be able to align your solar panels and antennas), - is to have the manned area only within the ring. small mechanical seals on conduits for ferrying liquids (which should be easier to make liquid tight than gas tight) and electric connections would be all you should need between the manned ring and the fixed unmanned section

an by placing this ring at one extremity of the spacecraft (only held on one side) you can have your docking hub being an integral part of the ring, in the middle (the docking hub + the descent spacecraft would be placed to be line up with the spine - the descent spacecraft would simply roll on itself in the middle) 

by keeping anything related to gases within the hab ring, you won't need any very large gas tight mechanical seals (as such seal would need to be large enough to let an astronaut go in  the middle) - you'll maybe only some liquid tight mechanical seals if you store water within the fixed section. 

of course, you would spin down the ring for docking / undocking the spacecraft - even if it is in the middle, docking with a rolling dock would be annoying ^^ (who says interstellar ? ;))

Edited February 3, 2016 by sgt_flyer

[SargeRho]

Wouldn't it make more sense to just spin the entire ship, and put the solar panels on a ring that stays still in that case?

[mikegarrison]

On 2/2/2016 at 0:17 AM, Porkjet said:

Been wondering about this problem for years. Nice thread, good answers!

Keeping a rotating section inside a static hull is what I've allways thought to be the most practical option too.

 

BTW I think that any spinning hab section be it internal or external or the whole ship should probably be stopped before attempting any manuvers, because of the gyroscopic effects keeping it stable along the rotation angle, kinda working against rotations along other axis. Have you ever held one of those powerballs in your hand that have a spinning weight inside that makes it very hard to rotate them? Same thing is probably the reason why bicycles dont fall over at speed or why some sattelites are spin-stabilised or why bullets are rotated by barrel thread. I guess this might be a problem for spaceships trying to maneuver.

FYI, gyroscopic forces have nothing to do with bicycle stability. Calculate the numbers and you'll see that bike wheels are much to light to contribute any significant gyroscopic forces to a bike. What keeps bikes stable is the geometry of the front axle location (and therefore the contact patch on the ground) relative to the angle of the head tube and where it projects to the ground. This is called "caster", and it's also why your car steering wheel will tend to center itself when you let go of it.

Edited February 4, 2016 by mikegarrison

[sgt_flyer]

@SargeRho

 

depends - that would make a lot more rotational inertia to stop when you want to spin down  - so that would be to check about tradeoffs  (as you'll need your fuel tanks for RCS (and maybe main engine too, if you don't want to spin down) to be able to work in both spinning (possible fuel centrifugation) and microgravity situations. having only the hab section rotating would limit the mass you need to stop when spinning down, and you only have to design your fuel tanks to work in for zero g - low linear acceleration situation (as there's no point to keep the spacecraft rotating when it's waiting in earth or mars orbit). 

having only the hab spinning could limit gyroscopic effects (as the mass and spin speed would be limited relatively to the fixed section) - so you could do more easily some manoeuvers without the need to spin down.

- so in the end, what would be the simpler solutions ?  more complex propulsion tankage, or more complex connection to the hab ring ?  (only to keep antennas and solar panels 'fixed')

Edited February 4, 2016 by sgt_flyer

[Nibb31]

8 hours ago, fredinno said:

Spinning the rings also gives you a energy storing option via flywheels. Also, you save propellant needed for constantly de-spinning every time you want to do a small burn.

It also loses energy through friction. And using it for energy storage means that you vary the RPM based on power requirements, which makes you have to choose between power draw and an optimal health environment. That's a bit counterproductive. I'm pretty sure that batteries with no moving parts are a more reliable and efficient way to store energy. 

Oh, and what happens to all that stored energy if a bearing in your rotating joint locks up and jams? It's not going to be pretty.

On the other hand, spinning the entire ship stabilizes the ship. No need for CMGs and less RCS propellant required. More weight saved.

Why do you insist on despinning for maneuvers? Any decent avionics computer should be able to maintain yaw and pitch while maintaining positive roll instead of zero roll. Even for docking/undocking it's no big deal as long as your docking ports are longitudinal.

Edited February 4, 2016 by Nibb31

[SomeGuy123]

54 minutes ago, Nibb31 said:

Why do you insist on despinning for maneuvers? Any decent avionics computer should be able to maintain yaw and pitch while maintaining positive roll instead of zero roll. Even for docking/undocking it's no big deal as long as your docking ports are longitudinal.

Physics.  If you draw out the actual vectors, and keep in mind that RCS thrusters cannot burn their fuel instantaneously, then you waste RCS fuel, always.  Only way to not waste fuel is if the burns take only microseconds to complete, and the gas isn't that fast to exit the thruster.  

Maybe if you change the thruster orientation during the burn keeping it aligned in a way that compensates for the rotation.

The reason your idea is still bad and will probably never be done in the history of manned spaceflight actually has to do with design coupling.  You talking about the complexity of electromagnetic bearings - yeah, that's some complexity - but you fail to appreciate that if you take a perfectly good rocket engine and fuel tank design and expect it to take rotational stresses that vary depending on the location on the ship - you end up with immensely complex engineering problems.  The current heat radiators and solar panels they fly are only able to work in microgravity, as a side note.

If the bearings don't have to hold an internal tube that astronauts can traverse, it's just a big bearing in vacuum, that's straightforward engineering.  

Edited February 4, 2016 by SomeGuy123

[Nibb31]

5 hours ago, SomeGuy123 said:

Physics.  If you draw out the actual vectors, and keep in mind that RCS thrusters cannot burn their fuel instantaneously, then you waste RCS fuel, always.  Only way to not waste fuel is if the burns take only microseconds to complete, and the gas isn't that fast to exit the thruster. 

That's nothing you can't engineer your way around. You could put your thrusters on Canfield joints as mentioned previously, or you can just use lots of "microsecond" burns, it will take longer, but we're talking about interplanetary timescales. Or you can accept to carry more propellant as part of the tradeoff of doing without all the complex joint/bearing/seals/motors/lubricants/radiators/spares associated with a rotation section.

Quote

The reason your idea is still bad and will probably never be done in the history of manned spaceflight actually has to do with design coupling.  You talking about the complexity of electromagnetic bearings - yeah, that's some complexity - but you fail to appreciate that if you take a perfectly good rocket engine and fuel tank design and expect it to take rotational stresses that vary depending on the location on the ship - you end up with immensely complex engineering problems.  The current heat radiators and solar panels they fly are only able to work in microgravity, as a side note.

You can't say that it will never be done.

The stresses aren't anything that can't be designed into the ship. There are plenty of engineering solutions to stress load problems. You could add spring loaded flexible hinges to take up a part of the load like on the Orion solar panels, as previously mentioned.

There are plenty of designs of spin stabilized spacecraft in the past and future concepts too. 

Remember ISEE-3 ? It's still operating after 30 years in space. Its thrusters are dead and fuel is depleted, but it's still spin stabilized, online, responding to commands, and its radial solar panels are still producing power. No moving parts.

 

Quote

If the bearings don't have to hold an internal tube that astronauts can traverse, it's just a big bearing in vacuum, that's straightforward engineering.  

Even if it's not pressurized, you're going to need connections between the axial structure and the rotating section. And making it unpressurized doesn't make things like lubrication or heat dissipation any easier. It also doesn't prevent a catastrophic lockup. Remember the SARJ failure on the ISS in 2007? They were able to fly up spare parts, but an interplanetary vehicle would have to carry spare bearings and replacement parts.

Edited February 4, 2016 by Nibb31