"2001 A Space Odyssey" space station: Rotating or contra-rotating?

[Miles Behind]

I assume that a space station like portrayed in 2001 A Space Odyssey could either be fixed in rotation, or contra-rotating. Is there an advantage of one over the other?

 

[DaveC426913]

What do you mean? How are you distinguishing rotating from contra rotating in a fixed, rigid structure like the wheel?

Are you suggesting each ring might rotate in opposite directions? That does not appear to be the case in the film's depiction, unless they make that change once the 2nd ring is finished.

And I believe there are gyroscopic problems with such a design.Or do you mean just the hub is contrarotating to remain fixed wrt the stars? This raises problems with getting and maintaining an airtight seal, which is non-trivial.

 

[sophiecentaur]

either be fixed in rotation, or contra-rotating.

The purpose of rotating would be to give an atificial g. I can't see there would be be advantage in not having the whole ship rotating. WHen docking, the cheapest solution would be to rotate the approaching ship.
Any measuring / navigation / comms / PV equipment which needs to be non-rotating relative to the local universe could be mounted on the ship's axis. Some kind of air lock would allow access for service crew.

 

[DaveC426913]

WHen docking, the cheapest solution would be to rotate the approaching ship.

Cheapest maybe, but not safest.

Matching rotation is fine but it is not very error tolerant. Small deviations can lead to big consequences.

 

[sophiecentaur]

Cheapest maybe, but not safest.

Matching rotation is fine but it is not very error tolerant. Small deviations can lead to big consequences.

But what difference would it make which one were rotating? We're in free space out here; everything is relative.

 

[DaveC426913]

But what difference would it make which one were rotating? We're in free space out here; everything is relative.

With both objects rotating: If the incoming ship moves off its docking axis by more than a little, it is at risk of colliding with the docking station walls as they rotate, different parts of which are moving at nontrivial velocity relative to the ship.

(Imagine the 2001 space clipper rotating on its axis to match the wheel's rotation so it can slide inside. They have zero relative velocity only if their axes are colinear.

If the clipper is coming in not dead centre, but, say, to the left by a few metres, it is still rotating on its own axis but that is no longer lined up with the station's axis.

If it continues on its course, the ship is at risk of getting clobbered by the docking station wall which could be rotating at quite a few m/s.

^{Picture is stock photo. It does not represent my scenario; it is merely a visual aid.}


Even if disaster is averted, the ship now has a much more complicated route to get back on track since it is still rotating its own axis . It may actually back off, realign and try again. )Whereas, if they are not rotating at all (relative or absolute), and the ship moves more than a little, the docking station is not at risk of clobbering the ship - the relative velocity will be trivial, since it is no more than the ship's own error - 1m/s. (Also, getting back into position is a trivial.)

IOW, in a nonrotating approach the consequences of an error are linear (1:1) - whereas in a rotating approach they can be arbitrarily multiplied (5?:1).

Everything is not all relative when rotation is involved.And as an added bonus, a nonrotating docking port's throughput is limited in capacity only by size. You could have a half dozen ships, all berthed side-by-side in their own bays, simultaneously and independently docking and undocking.

In a rotating port, exactly one ship can be in-play at a time, and must be moved into and out of the pipe to make way - a bottleneck. What if there's an accident in that bottleneck? Everyone's trapped.

 

[Filip Larsen]

Even if disaster is averted, the ship now has a much more complicated route to get back on track since it is still rotating its own axis .

Having the hub rotate with the station is by far the most simple solution in almost all respects except for docking, but this does not mean its impossible to make a fairly safe (automated) approach and docking/berthing mechanism (i.e. sensors/computers/actuators). Such mechanisms already exists today which take into account the relative orbital motion during approach and docking to a non-rotating station. To me it seems far more feasible to "extend" the capability of such technology to include spin-axis approaches than to design the whole bay to rotate relative to the station. Note that there will be a cylinder around the stations spin axis in which a spacecraft will be safe, i.e. not collide with the bay, so even if the approaching craft is a bit off-axis there is plenty of opportunity to safely nudge it back on-axis for a good berthing position (i.e. a position/attitude range where the spacecraft almost passively can maintain a near-zero relative motion while robotic arm grab the spacecraft).

Even for larger stations with larger bays it seems far more feasible to have the bay rotating (i.e. fixed with the rest of the station) and let any docking/berthing mechanisms (temporary) rotate, e.g. using rails or similar. For very large stations with very large cylindrical bays that potentially can house multiple spacecrafts its seems far more feasible to have the approaching spacecraft station themselves non-rotating over a section of the bay where robotic arm on a long rail (potentially going full circle around the bay) then can grab it and slowly accelerate it to zero relative motion with be bay. Such a mechanism also includes the possibility to do spin-axis docking with or without spacecraft rotation so it should be quite flexible (as a design baseline at least).

 

[DaveC426913]

Yes. A rotating bay may well be more efficient - as long as everything works as it should. My point was merely that it is not the best when it comes to safety margins. Faults are harder to recover from. Rotating docking has intrinsic relative velocities that become dangerous in a way nonrotating docks do not.

A five second lapse or glitch in attitude control at just the wrong moment can end up disastrous in a way that it cannot with a nonrotating dock. In other words, rotating is less fault-tolerant.

Naturally, as you point out, larger bays can have procedures that act to mitigate such risks - but that's for a future phase, innit?

 

[Filip Larsen]

A five second lapse or glitch in attitude control at just the wrong moment can end up disastrous in a way that it cannot with a nonrotating dock.

I don't see that. No matter if the station has a fixed or counter-rotating bay, approach procedures would required some safety margins outside which contact may indeed end with catastrophic results, but inside those margins there shouldn't be much difference. I guess my point is that a counter rotating bay in total is so much more complicated to design and has many more failure modes (also during docking) than a simple berthing arm in a fixed bay that can grab a near-axis spacecraft.

The fixed bay is just so much simpler. For instance, if the station spin axis is parallel to the orbital plane normal then an approaching spacecraft can "just park" near spin-axis just outside the bay and then completely passively let the relative orbit "push" it into the bay where it can be grabbed. In LEO that would take around 22 minutes and if the spacecraft CM is parked 20 m from station CM it would require a capture with residual speed of around 20 mm/s. If the bay is open all the way through a fail-safe procedure could be to simply not grab the craft, i.e. completely passive fail-safe.

Edit: Fixed spin axis direction from r-bar to h-bar in example above. Wording.

 

[sophiecentaur]

The fixed bay is just so much simpler.

I agree. The other possibilities that have been suggested still require that an additional dock rotates about the same axis as the main ship. In terms of the energy required to mate perfectly, surely that would be less for the smaller mass (i.e the visiting ship). The maths involved in accurate docking is pretty trivial (no harder than navigating towards the station in the first place. The visitor only needs to approach along the spin axis of the ship and then match the rotation.
I remember a past thread in which I suggested an approaching ship could make tangential contact and that idea , reasonably enough, was not well received. But I think that any space station would need artificial g so, however the docking is achieved the navigation problem is difficult. I'd imagine that the spin axis of the station should be parallel to the spin axis of the planet. Any thoughts on that?

 

[Rive]

To me it seems far more feasible to "extend" the capability of such technology to include spin-axis approaches than to design the whole bay to rotate relative to the station.

In some previous topic I had an idea about having rails around the outer edge of the rim, where the incoming ships can 'hang in' and 'speed up' till docking is possible.

 

[Filip Larsen]

[...] an approaching ship could make tangential contact

If the approaching (small) ship has to match rim speed before contact then this more or less corresponds to the SpaceX "tower pincher", only here the spacecraft arrives from "below" and has to be captured just as it reached closest approach. The trajectory of a spacecraft with matched rim speed as seen from the rim will locally look like parabolic up/down motion at the rim acceleration, so for a 1G rim this is fairly sudden load on the station. Alternatively, the capture point could be along one of spokes to go for, say, 0.5G only. In conjuction with capture the station also likely need to move some internal mass (e.g. water) to keep station GG near the spin axis. On release (i.e. the reverse operation) a proper timing can work as a full retro-burn if the timing relative to the entry point is correct, but if not the excess speed is probably going to be less useful. Since a tangential approach is a highly dynamic process it likely requires some fairly advanced automation, both during approach and capture in order to keep it both safe and reliable. One significant drawback of such a high-speed approach is that a "go-around" will likely be very fuel expensive as the spacecraft effectively will be on a rather different orbit than the station right before capture (i.e. with much higher apogee as an approach the "other way" likely will give a perigee resulting in reentry).

I'd imagine that the spin axis of the station should be parallel to the spin axis of the planet

The relative orbital motion will make approaches to a spinning station (whether is to the rim or center) rather difficult (at least timing-wise) if the station spin axis is not parallel to the station orbital plane normal. This will be then parallel to the planet spin axis only if the station orbit has zero inclination, i.e. orbits the planet equator.

 

[Filip Larsen]

the incoming ships can 'hang in' and 'speed up' till docking is possible

For a spinning station with spin axis parallel to its orbital plane normal, there is a line just before and after at the same orbital height (i.e. along v-bar) where an approaching spacecraft can approach as slow as needed and even hold position without spending any reaction mass. Any rail capture mechanism on the rim could then grab and accelerate the spacecraft as needed, but it also means the "underside" of the rim would be more or less fully dedicated to this mechanism (i.e. the rails) and its associated "safe zone". It also seems prudent to include in a docking design where the visiting mass is likely to end up at the station. It makes sense that humans would like to go to the rim, but cargo is much more manageable near the center so a rim approach, every thing else being equal, seems most appropriate for small passenger shuttles. Perhaps even only inter-orbit shuttles, e.g. small capsules that can ferry passengers and light cargo from a large spacecraft holding position near station to the station itself.

 

[Flyboy]

Imagine the 2001 space clipper rotating on its axis to match the wheel's rotation so it can slide inside.

Uhm… isn’t that what they did in the movie?

In regards to misalignment, if you’re off axis by more than a foot with modern guidance and navigation systems, you’re probably going to abort the docking anyway. And at 6RPM or less, you should have no trouble with making corrections for any misalignment. (6RPM is frequently quoted as the “nausea limit” for spin gravity, as anything more causes inner ear disruptions for most people.)

The assumption of having an actual docking bay also makes little sense, imo, because there’s little advantage to it. RCS plume impingement, radiators not working well, etc. Better to put your docking connection on the outside. If you absolutely have to have more than one docking port on a side, then use a manipulator system to capture the smaller ship at the axis, then swing/slide them outwards to the docking port. This would really only be suitable for use on really large stations, though, both from a physics/engineering standpoint and from a logistics standpoint.

 

[DaveC426913]

Uhm… isn’t that what they did in the movie

It is. But we try not to rely on film art directors for our space science.

 

[DaveC426913]

In regards to misalignment, if you’re off axis by more than a foot with modern guidance and navigation systems, you’re probably going to abort the docking anyway.

The example I mentioned was a systems failure at the wrong moment. Your retros dont fire or your electrical system glitches

Consider not just shuttles but other activity on a busy station such as repair crew in EVA suits. A crew person injures themselves or runs out of air or propellant.

To which you say: "You would not have crew making their own way into and out of the station on their own power; they would likely use a work shuttle to get to the work site."
To which I say: "Do you suppose there's any safety advantage to having crew be in a vehicle when entering and exiting the station?

 

[sophiecentaur]

so for a 1G rim this is fairly sudden load on the station. Alternatively, the capture point could be along one of spokes to go for, say, 0.5G only

It's the old ω²r problem. and that means you'd have to dock nearer the axis of the station than the radius of the torus. There would always be an impulse to deal with and spacecraft design is normally based on low stress structures.

To avoid any unbalanced impact the docking would need to be near the position of the CM which could call for two rotating tori. Sounds too much like hard work when the conventional axial approach is an almost 'trivial' problem.

 

[Flyboy]

The example I mentioned was a systems failure at the wrong moment. Your retros dont fire or your electrical system glitches

If you're disabled by that single failure to a point where you fail the docking, your design lacks sufficient redundancy for manned spaceflight, let alone docking.

Consider not just shuttles but other activity on a busy station such as repair crew in EVA suits. A crew person injures themselves or runs out of air or propellant.

A valid point on the EVA angle. Doing repairs via EVA on a spin station is a non-starter while it's spun up. You'll have to despin the station, use robotics, install extra infrastructure on the outside of the station to support astronauts on EVA and contain any FOD hazards from dropped tools/parts/etc.

To which you say: "You would not have crew making their own way into and out of the station on their own power; they would likely use a work shuttle to get to the work site."
To which I say: "Do you suppose there's any safety advantage to having crew be in a vehicle when entering and exiting the station?

I'm... not quite understanding what you're trying to say here.

It is. But we try not to rely on film art directors for our space science.

Valid. But I do appreciate it when they do their homework and use, if not accurate physics and science, then at least plausible. A minimum of handwavium is a good thing. Aside from the whole monoliths bit, 2001 is actually a surprisingly accurate hard science fiction setting.

 

[sophiecentaur]

. Doing repairs via EVA on a spin station is a non-starter while it's spun up

Why cast off when you could be tethered? A three point tether would give a very firm platform. The maintenance team would be like a spider on a web and they'd have familiar 'gravity' to help them.

 

[DaveC426913]

If you're disabled by that single failure to a point where you fail the docking, your design lacks sufficient redundancy for manned spaceflight, let alone docking.

Granted.

Doing repairs via EVA on a spin station is a non-starter while it's spun up

Do you think the (albeit fictional) crew spins down that wheel every time they need to do exterior repairs?

But I do appreciate it when they do their homework and use, if not accurate physics and science, then at least plausible

No one's suggesting implausible, just debating the merits and tradeoffs....and the maturity of the technology.
It is quite plausible that the tech simply can't support a reliable, airtight seal three hundred yards in circumference. A contrarotating hub is then dead-on-the-table, regardless of any potential safety advantages.

 

[Flyboy]

Granted.

Do you think the (albeit fictional) crew spins down that wheel every time they need to do exterior repairs? No one's suggesting implausible, just debating the merits and tradeoffs....and the maturity of the technology.
It is quite plausible that the tech simply can't support a reliable, airtight seal three hundred yards in circumference. A contrarotating hub is then dead-on-the-table, regardless of any potential safety advantages.

What studies I have seen for spin gravity have usually involved smaller scale designs that are deployable, and the seal between rotating and stationary sections is always an issue even at ~1-2m diameters. I’ll have to dig around a bit tomorrow, see if I can find the report I’m thinking of. I seem to recall it being part of the Nautilus-X proposal from NASA?

 

[Rive]

Doing repairs via EVA on a spin station is a non-starter while it's spun up.

Those rails could make that trivial too.
These rail-based 'travelers' are permanently installed on many bridges.

 

[sophiecentaur]

Those rails could make that trivial too.

The term 'rails' implies a permanent structure. The same effect can be achieved in the presence of artificial g with three tethers, forming a tetrahedron with vertices at fixed points on the hull. A structure with artificial g would need (perhaps) to be stronger than a non-spinning hull and that would apply to any tether system but there would be very little fuel needed for manouvering around of maintenance craft too. Very fail safe too.

 

[Rive]

The same effect can be achieved in the presence of artificial g with three tethers, forming a tetrahedron with vertices at fixed points on the hull.

Sure, but that would not work as guide for ring-docking (that's where the rail came from).

 

[sophiecentaur]

I realise that EVA has been introduced as a possible design constraint for a spinning station. A ring docking system was suggested in the context of EVA (above). But docking to a tether could also be arranged. Angular momentum needs to be considered for any docking system for a rotating station.
Ring docking is not as straightforward as suggested because of the 1g (or more) impulsive radial force at the time of attachment. Some sort of bungee connection would be needed to remove this and this would be more along the lines of a tether than just latching onto a rail.
I have yet to be convinced that there are any inherent advantages over an axial approach, except for the need for one at a time docking at each end of the axis. Heathrow Airport seems to work fine with very heavy traffic and only a single runway so traffic control would be no problem using axial docking.

 

[Filip Larsen]

Regarding the orbit of a spacecraft needed for making a rim capture (i.e. matching the rim speed) on a spinning station I earlier said:

One significant drawback of such a high-speed approach is that a "go-around" will likely be very fuel expensive as the spacecraft effectively will be on a rather different orbit than the station right before capture

Now I got around to model this a bit more I think I have to retract that statement.

Assuming the spinning station has a circular orbit of radius $r$, then if the spacecraft is placed in an elliptic orbit (eccentricity $e > 0$) with semi-major axis equal to $r$ then the spacecraft will be in an orbit with same period as the station and therefore the two will (ignoring perturbations from non-spherical Earth and similar) pass near each other again one orbit later, which means a new capture can be retried with "minimal" fuel use. In LEO, a "go around" would thus take around 90 min.

Such an orbit also means the orbital speed of the spacecraft is same as the stations orbital speed $v_r$ near capture, but the velocity direction will be slightly different. Since the flight path angle $\gamma$ (which is zero for horizontal velocity) near capture can be related to eccentricity as $$\cos^2\gamma=1-e^2$$ and relative speed between the station and the spacecraft near capture is $$v_s = 2v_r\sin(\tfrac{1}{2}\gamma),$$ it follows after some calculations that
$$v_s^2 = a_s s = 2 v_r^2(1-\sqrt{1-e^2}),$$ where $a_s$ is the rim acceleration, $v_s$ the rim speed and $s$ is the rim radius. Since in LEO there is a lower limit to the perigee distance $r_p$ (to stay well away from reentry) one can use $r_p = r(1-e)$ and $v_r^2 = \frac{\mu}{r}$ to further get $$a_s s \le \frac{2\mu}{r}(1-\frac{\sqrt{(2r - r_p)r_p}}{r}).$$

For example, a station at orbital height of 300 km with rim acceleration of 1G and with perigee height of 200km, rim radius has to be less than around 1365 m which is fairly large. Thus it seems that the approach orbit itself is probably is not that critical as I thought in terms of reentry and go-around limits, even for fairly high rim speeds and fairly low orbit stations.

Edit: Fixed wrong statement about 2 near passes between spacecraft and station per orbit, since they of course only pass near each other once per orbit. The other time where spacecraft and station is at same orbital height the spacecraft will be far behind the station.

 

[sophiecentaur]

but the velocity direction will be slightly different.

Isn't that the real problem? The arrival can be given any orbit that you choose but the sudden change of direction has to involve a significant burst of thrust (or a jolt from a bungee arrangement). An axial docking can be performed at leisure and is more or less fail safe, compared with a 'grab it now' operation.

Transferring into a Moon orbit from am Earth orbit is much less of a problem than this form of docking because the Moon uses regular orbital dynamics (inverse square law) but a rotating space station cannot easily provide an artificial version of that.

 

[Filip Larsen]

The arrival can be given any orbit that you choose but the sudden change of direction has to involve a significant burst of thrust

If the difference in velocity exactly matches the rim speed right at the capture point on the rim then there is no thrust involved, which is the setup I tried to model in my post. The spaceship would be free falling until right at capture where it would then experience the rim acceleration. Its a bit like being catapulted into the air on Earth with a precise speed only to have a platform slide in under your feet just as you reach the top point. Or imaging falling upwards right next to a ladder you grab it the just at the moment you top out. Highly dynamics situation but not physically impossible.

An axial docking can be performed at leisure and is more or less fail safe, compared with a 'grab it now' operation.

I fully agree. I was just qualifying my earlier statement with some math showing that the range of "useful" approach orbits for a rim capture perhaps not are so narrowly limited as I originally thought. A rim capture would likely still require significantly more advance technology and yet, as I see it, provide no benefit over a slow controlled spin axis approach and capture which is almost feasible with our current technology.

However, even if rim capture seems a very unrealistic possibility, the reverse might be interesting though, i.e. releasing objects from the rim. For example, "escape pods" may be able to reach partial or, depending on the rim speed, even full reentry orbit simply by releasing at the proper time from the rim. On the other hand, this also means someone falling off at the wrong time may find themselves in a lot of trouble or at least a bit more trouble than the otherwise quite serious trouble of falling of a spinning station in general.

 

[sophiecentaur]

If the difference in velocity exactly matches the rim speed right at the capture point on the rim then there is no thrust involved,

Don't forget that a 'sharp left hand turn' as the ship is grabbed onto the rail involves a step change of 1G force on its structure as it starts to follow a circular path. That would mean the ship needs to be built with the hull strength to carry that 'weight'. That would mean the possible payload for the arriving ship would be less. A zero G design for axial docking would eliminate this problem.
All ships which are built or operated under fractional g conditions could have a slender design and save on materials. This would make them incompatible.

For example, "escape pods"

I like that idea. An escape pod, released from the outer edge of the station on a tether that's paid out gradually could be sling-shot at several times the circumferential velocity, using no fuel.

 

[Nik_2213]

Tangential, the many issues listed above drove me to a 'saddle-tank' arrangement for my Convention tales' 'City-Class' star-ships. They're very modular, with boxy hexagonal flanges/lozenges spaced by triple truss girders. Think stack of three-legged, 'Jack-up Barges'. Each flange carries three 150 metre nominal out-rigger legs with ports at 25 metre nominal spacing. A single cargo, tankage or spin-drum pod may be fitted on-axis between the triple spines, or 'several' docked lengthwise between successive flanges' leg-pairs.
A standard pod is 100_L x 50_D metres, so three may fit between successive leg-pairs, up to nine per flange pair. A basic spin-drum's carousel is ~47_D metres internally due to stationary 'guard skin'. If pod is multi-use, the carousel may be much shorter. Still, as ~47_D metres is a tad narrow for comfort at even ¼ g 'whirl', often taking several days to get your 'spin legs', many ships carry 'Super-Drums', 100_L x 100_D metres or, via adaptors, 200_L x 100 _D metres. Latter may be standard aboard 'future 'Capital Class' ships...
Shuttles dock into un-spun 'Holds', or to vacant ports / side-ports on flange 'legs'. Compared to the alternatives, moving from stationary leg to stationary pod to pod's spin-drum, even with baggage, is much, much easier done 'on-axis in full atmosphere'...

 

[sophiecentaur]

They have zero relative velocity only if their axes are colinear.

Jow is this any more of a problem than an aircraft landing on a runway that's not aligned with the wind (which is most of the time). Planes use suspension and tyres to cope with far worse situation than would exist in a spacecraft docking at almost zero velocity.

But we try not to rely on film art directors for our space science.

This is a perennial problem with space fiction discussion when real Engineering or Science are introduced. It's strange that recent SciFi movies are seldom quoted when we try to discuss what things would really be like. We always draw on 2001 for our sources.

 

[Ibix]

However, even if rim capture seems a very unrealistic possibility, the reverse might be interesting though, i.e. releasing objects from the rim.

Look up Babylon 5's Cobra drop bays. The series used this exact approach to launch Starfury fighters from the rim of the eponymous O'Neil cylinder by just opening a door under the fighters and letting them go. The visuals have a lot of artistic licence - the fighters somehow come away with a radial velocity instead of a tangential one - but the idea was there.

 

[sbrothy]

Look up Babylon 5's Cobra drop bays. The series used this exact approach to launch Starfury fighters from the rim of the eponymous O'Neil cylinder by just opening a door under the fighters and letting them go. The visuals have a lot of artistic licence - the fighters somehow come away with a radial velocity instead of a tangential one - but the idea was there.

Atomic Rockets argues pretty hard that space fighters make little tactical or economical sense:

"I know all you Battlestar Galactica fans are not going to want to hear it, but looking from a cost/benefit analysis, space fighter craft do not make any sense. Go to the Future War Stories blog and read the post Future War Stories".

Ofcourse we all want them to be there because they're cool and fit exactly one protagonist.

 

[DaveC426913]

Jow is this any more of a problem than an aircraft landing on a runway that's not aligned with the wind (which is most of the time). Planes use suspension and tyres to cope with far worse situation than would exist in a spacecraft docking at almost zero velocity.

Just so we're clear, this is the scenario I'm describing.

No amount of suspension and heavy-duty tires will minimize a broadside collision with a station wall.

 

[sophiecentaur]

No amount of suspension and heavy-duty tires will minimize a broadside collision with a station wall.

It's a bit 'how long is a piece of string?' without some idea of the actual speeds involved. A plane will be landing at 100mph and can't go much slower. A docking space ship can be approaching at less than walking speed and can slow down, reverse or nudge left and right. This is more like the mooring a boat scenario. We use inflated fenders for a 4ton boat coming up against a dockside and they work fine for dealing with a nudge of 1knot. I really can't see any significant problem with this aspect of space navigation.

Your images assume a certain style of engineering - to make good cinema pictures. After a long space journey it would be no hardship to approach your destination over several hours; no need to zooooom in to the dock.