Space station artificial gravity - how to spin up to speed?

[SteveO33]

Like the topic, and the efforts to solve the 2001: Space Odyssey designs.. However, if you're trying to help the crew, there may be an opportunity to just rotate the crew quarters and still achieve 95% of the goal - better crew health - with much less complication. The working areas, docking and the rest of the station would remain non-rotating and zero-g. This would reduce significantly the issues with balance/rebalance, mass requirements, docking, etc. Then you could start working with the medical guys to determine what minimum percentage of gravity would be acceptable. If they come back with say 50%, now you may have managed the design into something that can be achieved. I could envision two crew quarters powered by the station, rotating in opposite directions with fluid dynamic balancers for when crew entered and left to provide a stable, 50% G environment. The quarters would support dining, sleep, shower, toilet, rest, exercise, relaxation, leisure - life - at some level of gravity. Work would still be at zero-g.

 

[sophiecentaur]

Where is the point of the rail then?

You have to have something to keep the ship from carrying on along it tangential path. What's the alternative - magnetic wheels? The rail would be supporting the weight of the ship (or whatever fraction that the local g happens to be).
Are you sure you have my design sussed? Some of your questions and comments are worrying me.

 

[DaveC426913]

You really have still failed to grasp what I am describing. The Rail is part of the periphery (which is never 'at rest' relative to the CM because it's spinning around the CM) and would be fixed along the outer circumference. How would its speed be other than the same as the ship when it lands.

I'm not sure I see your model correctly either.

Let's clarify. To dock at the rim of a rotating station, a shuttle will be moving relative to some part of the space station at a not-insignificant speed. There's two ways to do it, I'm not sure which one you propose.

Just before contact, what is the shuttle's velocity wrt to station proper? Zero? Or positive?
Just before contact, what is the shuttle's velocity wrt to station rim? Zero? Or positive?

You have to have something to keep the ship from carrying on along it tangential path.

At the moment of contact, what is the relative velocity of the shuttle and rail?

 

[sophiecentaur]

At the moment of contact, what is the relative velocity of the shuttle and rail?

Zero. It aims to contact the rail (outer circumference of the wheel - plus an additional distance) at zero relative speed. Actually, just the same speed that an aircraft aims at the ground when it lands. (but it needs air speed to keep flying) so let's say helicopter. The relative velocities of the contact point to all other points on the wheel is non zero. I really can't see how it is a problem to arrange the approach of the ship to coincide in time and space and velocity, with a part of the periphery. Having a rail around the whole wheel. there will always be a point that the ship can attatch to. As it's a continuous rail, the ship can even have a bit of forward speed and ti will still be captured satisfactorily. A small amount of (mechanical) braking will take care of that.
My main concern now is to get across the actual idea of the docking procedure. Once raw system has been understood I am open for well founded objections to it.
1. Would it work?
2. Is it worth the effort"
3. What could be the advantages?

I acknowledge that it would be a totally different approach from what's done at present. Objections 'as a matter of principle' or due to misunderstanding are not helpful.
It strikes me that any system that involves a zero velocity connection would be worth while considering.

https://www.physicsforums.com/threads/space-station-artificial-gravity-how-to-spin-up-to-speed.900043/members/steveo33.613758/ 's comments are very relevant here - restrict the regions of non zero gravity to R and R for the crew. It's true that many reasons for space experiments are associated with a microgravity environment. I guess it would depend on the basic purpose that the station is up there for.
Reading this thread only confirms me in the view that it's would be a pretty naff job for most people.

 

[mfb]

Zero. It aims to contact the rail (outer circumference of the wheel - plus an additional distance) at zero relative speed.

Where is the point in wheels or rails then?

1. Would it work?

If you want to risk the integrity of the space station with every single approach: sure.

2. Is it worth the effort"

I still don't see advantages. I see many disadvantages.

Zero-g areas can be arbitrarily large - connected to the rotating part (living quarters, hotels, whatever) at the center.

 

[sophiecentaur]

The wheel and rail arrangement will allow for a small difference in speed and, after connection, allow the ship to move around to a suitable port to enter the station.
Considering that, with the present docking system, you have a few tens of cm of location accuracy and a very small speed tolerance and with the peripheral approach you would have as much as your tethering / hooking system design needed and that your relative connecting velocity could have an accuracy of several m/s, the landing tolerances seem pretty slack to me. There would be nothing in the ship's path that it would be approaching at high speed during the attachment process. Of course, it would be going relatively fast, relative to the the station CM - but then, everything else is too.

 

[DaveC426913]

With your system, your shuttle must be moving relative to the station at a not-insignificant velocity. That means dangerous KE is associated with every single docking.

It also means a huge inefficiency if the docking must be aborted. They have to completely reverse course, move back to their original position and regain the original velocity.

Docking at the hub is, by comparsion, a leisurely affair.

 

[A.T.]

Not sure if it was mentioned: Docking at the rim would shift the centre of mass more, and thus create more wobble.

Yes. That's true but so will re arranging the cargo around the station (particularly) during construction. So some sort of shifting counterweight would be required in any case.

Moving cargo around can be done in small bits, as slowly and carefully as needed. Docking a massive spaceship to the circumference is quite different.

It also means a huge inefficiency if the docking must be aborted. They have to completely reverse course, move back to their original position and regain the original velocity.

That's the main problem. And to make it worse, it also makes a docking fail more likely, by limiting the available time. At the hub you can float nearby without using fuel until you fix some problem.

 

[sophiecentaur]

your shuttle must be moving relative to the station at a not-insignificant velocity.

Relative to the CM, yes but every part of the station is moving relative to the CM. How relevant is that? The speed relative to the contact point would be nearly zero. Also, the rails and a possible tether / arrestor wire system would mean that the approach parameters could actually be fairly lax.
An aborted approach would involve carrying on and then making an approach from the 'downstream' direction. That would be to a diametrically opposite position on the wheel. The required impulse would correspond to a change of 60m/s. Is that a lot, in the context of all the other fuel used in getting there?
Docking at the hub does actually have some disadvantages. There can be only two docking positions (one each side), as they have to be sited exactly on the axis. That restricts the number of possible visitors and also poses a problem with a malfunction. However big the station happens to be, this restriction applies. Not so if the station is hybrid g-non g, of course but each dock does require the full capture/dock facility, where the two functions would be separate with the peripheral approach.The Starship Enterprise has a really easy life, in this respect.

 

[sophiecentaur]

That's the main problem. And to make it worse, it also makes a docking fail more likely, by limiting the available time. At the hub you can float nearby without using fuel until you fix some problem.

Hanging around so you can fix a problem is a very relevant advantage, I agree but the 'time' constraint can be offset against the position and speed constraints, which can be much less stringent. With a continuous rail, around the periphery, the time you arrive doesn't need to be specified. All that's necessary is to be able to grab your chance as your ship passes through the 'possible capture zone'. All this is so reminiscent for me when I think of the mooring and docking procedure of a medium to large boat. If you are all ready with your warps, you can easily pick an optimum time to make your connection with land or mooring. No highly accurate forward time calculation is required as you just pick the turning value in the procedure.
The choice of which system to use could depend on the actual size of the station. You could imagine that there could be congestion with comings and goings of several ships at a time on a large station.

 

[A.T.]

...position and speed constraints, which can be much less stringent.

I don't see that at all. For the same size of the "initial connector" you have the same position tolerance. For the same robustness of the "initial connector" you have the same relative speed tolerance. But at the hub you have far more time to make the connection, and the "initial connector" doesn't have to be robust enough to provide the centripetal force to the ship.

 

[mfb]

The required impulse would correspond to a change of 60m/s. Is that a lot, in the context of all the other fuel used in getting there?

Yes. 60 m/s is a significant orbital maneuver. The Space Shuttle had a total delta_v of 300 m/s, Dragon V2 will have similar values. Most of this is used for approaching the station and de-orbiting. Adding 60 m/s or even 120 m/s contingency for docking aborts is significant.

Docking at the hub does actually have some disadvantages. There can be only two docking positions (one each side), as they have to be sited exactly on the axis.

They don't have to be. The docking port does not have to rotate and can be as large as necessary. It can be connected to the rotating part at the center. Assuming the space station will be used for research, you want a zero-g section anyway.

Every position tolerance that a docking system at the rim can provide can also be provided by docking at the center. But a docking system at the center does not need large tolerances, which makes docking systems easier. It also allows standardized and androgynous docking systems. A rail docking system would mean spacecraft s could only dock to space stations, but not to other spacecraft s.

of course but each dock does require the full capture/dock facility, where the two functions would be separate with the peripheral approach.

You need a docking port for every spacecraft connected at the same time, that is the same for every approach.

 

[sophiecentaur]

I can see I am going to have to yield on this one chaps. It has been a very interesting exercise. The clincher was when I realized that the station need not be totally rotating.
I am possibly in a minority (too realistic, I guess) that doesn't reckon as space flight (let alone space living) as being my ultimate possible experience. I just can't imagine people wanting to spend extended periods in a space station (nor in a starship) and I can't help feeling that a reasonable level of g would make life much more tolerable and that only micro g experiments would be conducted in micro g conditions. But you have to ask what the purpose of the station would be. I have read the view that the main reason for long term space experiments is that they can use the micro g available but present day astronauts are subjected to enormously debilitating effects of long term micro g.
It's an interesting scenario to discuss though.

You need a docking port for every spacecraft connected at the same time, that is the same for every approach.

Actually, you only need one access port each. Getting attached can be a separate function. Airports have many more gates than runways and that was my reasoning. But it would all depend on the scale of the station as to the level of traffic involved.

 

[DaveC426913]

Relative to the CM, yes but every part of the station is moving relative to the CM. How relevant is that? The speed relative to the contact point would be nearly zero.

For an instant. The window for contact is constrained very tightly by space as well as time.

This is a recipe for disaster.

 

[Baluncore]

The necessary technology to spin up a space station has been available for over a century. Make a big Crookes Radiometer by fitting paddles to the periphery of the station. Paint one side white and the other side black so solar radiation pressure can spin up the station. Then rotate half of the paddles to face backwards which will regulate the speed. Quick adjustments to rotation rate can be made by hanging stores from the hub inside the spokes. Changing the length of the winch wire will change the radius to the centre of mass of the stores and so will change rotational velocity.

 

[jbriggs444]

The necessary technology to spin up a space station has been available for over a century. Make a big Crookes Radiometer by fitting paddles to the periphery of the station. Paint one side white and the other side black so solar radiation pressure can spin up the station. Then rotate half of the paddles to face backwards which will regulate the speed. Quick adjustments to rotation rate can be made by hanging stores from the hub inside the spokes. Changing the length of the winch wire will change the radius to the centre of mass of the stores and so will change rotational velocity.

But what to do with the gerbils already hired to keep the station turning?

 

[Baluncore]

But what to do with the gerbils already hired to keep the station turning?

Don't feed them, eat them.

 

[anorlunda]

It doesn't make sense to compare an electricity source to propellant. What are you going to do with electricity? Accelerate propellant. That's the ion thruster case I calculated.

Yes it does make sense. You could use solar to power an electric traction motor to accelerate rotation of a reaction wheel. No propellant needed. I think that was the thrust of the OP's question.

 

[mfb]

The necessary technology to spin up a space station has been available for over a century. Make a big Crookes Radiometer by fitting paddles to the periphery of the station. Paint one side white and the other side black so solar radiation pressure can spin up the station. Then rotate half of the paddles to face backwards which will regulate the speed. Quick adjustments to rotation rate can be made by hanging stores from the hub inside the spokes. Changing the length of the winch wire will change the radius to the centre of mass of the stores and so will change rotational velocity.

Radiation pressure is 3 μPa. For a very lightweight station of 100 kg/m², and assuming perfect heat conduction away from the black side, you get an acceleration of 3*10^-8 m/s² at the place of your paddles. Spinning the station up to 30 m/s with radiation pressure will take 30 years.
No way.

Yes it does make sense. You could use solar to power an electric traction motor to accelerate rotation of a reaction wheel. No propellant needed. I think that was the thrust of the OP's question.

That was an alternative discussed earlier, yes, but not in the post you quoted.

 

[Al_]

The necessary technology to spin up a space station has been available for over a century. Make a big Crookes Radiometer by fitting paddles to the periphery of the station. Paint one side white and the other side black so solar radiation pressure can spin up the station. Then rotate half of the paddles to face backwards which will regulate the speed. Quick adjustments to rotation rate can be made by hanging stores from the hub inside the spokes. Changing the length of the winch wire will change the radius to the centre of mass of the stores and so will change rotational velocity.

Radiation pressure will work very, very slowly, to start with.
Moving masses away from the CM can't slow the rotation to a halt completely. Space inside the station is at a premium so unlikely to be possible to move lots of mass to near the rim.
It might not even have spokes, it might have low-g upper floors.
The arrangement doesn't cancel gyroscopic forces.

 

[sophiecentaur]

. The window for contact is constrained very tightly by space as well as time.

Not really. The actual time that your path takes you to coincide with the (continuous) periphery rail hardly matters at all if the ship can make contact with any point on the perimeter rail. A boat arriving at an uncluttered mooring pontoon only needs to get its course and speed right and can arrange to touch on at any time.
In the sort of system I am thinking of, the capture mechanism could extend over several metres of intercept area. I wouldn't call that a tight constraint, considering the last very successful landing system that was used on Mars, in which timing was highly relevant. (Who would have thought of sky crane without knowing a lot about space landings?)
The consequence of a flyby due to 'failure to connect' would be, say, 60m/s worth of fuel and not any sort of disaster; just an expensive one-off. An actual collision with an inner part of the wheel would not be good news but that would be predicted with loads of time to spare and the abort procedure would be very similar to that for any medium speed collision.
I am no longer touting my system for business but I want to make sure that the objections are not just intuitive. People seem to be ignoring the fact that, in space manoeuvres, there are no surprises due to wind and waves. Also, we are not talking in terms of a man with a joystick and a thrust control for this sort of system.
Any capture mechanism could be expected to cope with a (radial) distance offset of a metre or so and I can't see that the navigation would have a problem with flying the ship to an accuracy of course of a fraction of a metre. The actual time of the contact is not very significant; what counts is the distance between the rail and ship's course.
Aiming to kiss the rail but ending up 1m closer will result in a radial ('vertical')speed of around 4m/s. 0.1 m gets a vertical speed of around 1m/s. How does that compare with aircraft undercarriage performance? Reading various flying forums, I get the impression that a commercial jet makes a 'good' landing at around 1m/s and an undercarriage can cope with several times that. There is no direct comparison between the two structures and an aircraft undercarriage has other things to deal with but the cost of a bad aircraft landing can be the deaths of hundreds of passengers, which is comparable with the disaster of taking our a whole space station. In either case, it needs to be 'right'. But the 'landing would be arranged to be appreciably outside the surface with a very low possibility of 'ploughing in'. I would envisage the ship, once captured, would be (in their view) hanging from the rail by an appropriate length of suspension. Once settled, it would move along the rail and branch off to an entry port,. Another ship could arrive or depart very soon after - just like at an ariport..
Those figures are not sufficient to reject the idea - even if there are lots of other valid objections to the system. My scenario is based around a busy and pretty massive space station. That could well make it a totally unreasonable suggestion - but that's not the purpose of this post.

 

[mfb]

The actual time that your path takes you to coincide with the (continuous) periphery rail hardly matters at all if the ship can make contact with any point on the perimeter rail.

10 seconds before docking, you know you'll either dock in 10 seconds (+- a very small window for steering) or not at all in this attempt. If something needs more time, you have to scrub the attempt, waste 60 m/s of delta_v and try again.

That's an issue you don't have if you dock to the non-rotating part.

 

[sophiecentaur]

10 seconds before docking, you know you'll either dock in 10 seconds (+- a very small window for steering) or not at all in this attempt. If something needs more time, you have to scrub the attempt, waste 60 m/s of delta_v and try again.
That's an issue you don't have if you dock to the non-rotating part.

I should hope that you would know if your course was good along a tangent a lot more than 10s before contact. I'd reckon in terms of many minutes, with plenty of time to correct course. Speed difference could be +/- 0.1m/s which would be well within the capabilities of a coupling mechanism. You could do that by hand and eye, even.
If you got it a bit wrong and you kiss at 9.5 seconds, because of a small difference in velocity, what is the difference? You latch on to a different part of the rail. The count down will be useful for the system to know when to deploy its grab / net / hook but why do you say you need to plan your rendezvous at a particular time? You draw a line on your nav chart (virtual) and, if that line is a tangent, that's all you need. Yes, you could miss but what could cause that miss? Is there a wind or a tide or waves, or pockets of warm air / turbulence? You must be subconsciously basing a lot of your objection on those things. There are none. Just consider the accuracy with which slingshot orbits involve and they are not something you can do much correction for. Fine tuning yes but you have only one shot at them and they have far more imponderables. Ship and station are comparatively so near to each other, when the docking course is set that the calculations have to be relatively straightforward.
What "something" are you suggesting could need more time? The tolerance to errors is way higher than you imply. Time doesn't matter. Speed difference is nothing like as important compared with a head-on docking. Course is important but you still have a up to a metre. according to my sums. (I do wish more people would commit themselves to some calculations in situations like this one. I am always grateful when someone shows me an error on my sums.)
I have already yielded to the point that docking with a non-rotating part has fewer obvious problems. I am just setting out some hard figures relating to the peripheral docking because the instinctive reactions do not agree with the simple sums.

 

[mfb]

Let me rephrase it:

300 meters before docking, you know you'll either dock in 10 seconds (+- a very small window for steering) or not at all in this attempt. If something needs more time, you have to scrub the attempt, waste 60 m/s of delta_v and try again.

300 meters. The time where currently a docking process still takes more than one hour. Why? To make sure the position and velocity are correct and can be controlled in every step, with multiple course corrections within this hour.

If you got it a bit wrong and you kiss at 9.5 seconds, because of a small difference in velocity, what is the difference?

You'll be off by 15 meter in some random direction if your velocity vector is wrong in arbitray directions.

 

[A.T.]

If you got it a bit wrong and you kiss at 9.5 seconds, because of a small difference in velocity, what is the difference? You latch on to a different part of the rail.

It's not about arriving late for dinner, but about having enough time correct the course. Velocity is a vector that can be wrong in magnitude and direction.

 

[sophiecentaur]

Let me rephrase it, too.
As long as your path is tangential and your speed is within fairly lax limits, the actual time of your arrival doesn't matter. I still get the feeling that you have not understood what I am describing. So many of your questions are based on colliding with the centre of the wheel. The centre of the wheel ( and the docking bay of the ISS) is a 'solid' object with little resilience. What I am discussing is nothing like that; it doesn't need to be because contact is 'outside' the station body. It can be flexible and allow for much more variation in position and speed. If you still have a problem, try drawing out the situation and do the same sums I have done, to work out the relative velocities and positions. The mechanical requirements are very much on the same lines as for aircraft landing gear. You do not seem to take that into account.
If you can tell me of a mechanism that can stop the ship from traveling in a straight line whilst drifting towards the station and how that line cannot be set up at a significant distance away then I could take on board what you are claiming.
I understand that the tolerances (lateral, alignment and speed) are very stringent for a normal docking. But the situation is really not the same for what I am describing. Why do you keep insisting that it is?

You'll be off by 15 meter in some random direction if your velocity vector is wrong in arbitray directions.

Why should it be, if all you are doing is aiming to keep to a line that's a tangent to the wheel? Do you not see how your time of arrival is of minor importance?

 

[A.T.]

all you are doing is aiming to keep to a line that's a tangent to the wheel

Which requires corrections for which you have limited time, because you have to approach the point of tangency at a fixed speed. At the hub you can approach as slowly as you wish.

 

[Baluncore]

Radiation pressure will work very, very slowly, to start with.

Yes. But I wanted to reapply riverboat paddle-wheel technology to the future.

I was also considering delivery of construction material by attachment while passing on the fly to a spinning loop, a polygon, or a star of carbon fibre cable with a great radius. Then using solar power to operate winches that symmetrically pull the construction modules towards the centre with the radius reduction and inherent angular acceleration.

It might not even have spokes, it might have low-g upper floors.

I think a docking hub with two spokes would be essential for access and construction. Note that transport of stores from the hub to the periphery would decelerate the 'station'.

The access hub of a rotating station is another possible application for the anti-twister mechanism.
https://en.wikipedia.org/wiki/Anti-twister_mechanism

 

[sophiecentaur]

Which requires corrections for which you have limited time, because you have to approach the point of tangency at a fixed speed. At the hub you can approach as slowly as you wish.

I have asked, in vain for some numbers here. How is it that the best artillery can achieve such high accuracy in placing shots under all the perturbations you get on Earth but you guys are claiming that course corrections are constantly necessary in the space docking scenario? What is it that can disturb the relative velocity of the ship so that a tangent to the wheel will be waving about? I realize that, over a longish period, there will be gravitational effects but will there really affect the motion, once the ship is a few km from the station? The objections I read have been a bit too arm waving and not quoting calculations or results of experiment. I am totally prepared to be convinced but the least you guys can do is to give me some numbers. PF threads don't work well on 'assurances' alone.
You can tell me as often as you like that a near zero velocity approach to the hub has its advantages. I accept that.

 

[Baluncore]

I don't think it is necessary to dock a delivery vessel to the rotating station to unload materials.

Delivery of most of the stores, material and crew to the rotating station should be by attachment of a modular delivery capsule to a long station tether while on a synchronous tangential flyby. The delivery capsule is pulled in, unloaded, then dismantled for reuse. The capsule would be assembled from a regular set of parts that are reused to fabricate more station modules.

Passengers in transit or crew on exchange could arrive and depart from the hub so as to save materials and conserve rotational energy. Exports should depart from the central hub. With time the station can grow in size while maintaining the same virtual gravity.

 

[A.T.]

I have asked, in vain for some numbers here.

https://en.wikipedia.org/wiki/Space_rendezvous#Rendezvous_phases

They currently have 45-90min for corrections over the last 100m. With your method they would have only a few seconds.

 

[sophiecentaur]

https://en.wikipedia.org/wiki/Space_rendezvous#Rendezvous_phases

They currently have 45-90min for corrections over the last 100m. With your method they would have only a few seconds.

Can you explain that reasoning? That argument would suggest that landing an aircraft would be too risky to undertake.
The fact that present docking uses a long period of time is no reason to suggest that the same limits apply to the alternative method. Position, velocity and time are all less critical. You do not need to have zero relative speed and the lateral tolerance could be a matter of metres. Please tell me what perturbations can be present to cause a problem. Does the 30m/s speed difference produce an incalculable uncertainty in the rendezvous point?

 

[A.T.]

That argument would suggest that landing an aircraft would be too risky to undertake.

An aircraft cannot hover / land slowly without using lots of fuel and be specially build for it. A spaceship can do this naturally.

 

[sophiecentaur]

An aircraft cannot hover / land slowly without using lots of fuel and be specially build for it. A spaceship can do this naturally.

What is the relevance to that remark? Aircraft land perfectly well and, but for the vagaries of weather and the military, they would have an even better record of making contact with Earth at high speed. If an aircraft could hover for free, are you sure it would wait for 45-90 minutes to land, as a spacecraft takes to dock? You are not putting logical arguments together here. I asked for some (quantitative) reason to suggest that the accuracy of manoeuvring would be as bad as you claim. If it works for aircraft the why not for spacecraft ? (Same potential cost of human life.)

 

[mfb]

Aircraft don't land on ceilings where they have to get attached within a fraction of a second. They are also quite sturdy - their mass is not as important as it is for spacecraft , so they can handle larger relative velocities and accelerations. Compare to spacecraft s hey have a very large region to land, even on aircraft carriers.

Aircraft that can land vertically do that slowly as well. They don't take an hour for that, but they also don't do a hoverslam where you reach zero velocity exactly on contact.