# How can you make a vacuum proof rotatable seal?

Many systems have been discussed where you have a rotating spacecraft for artificial gravity purposes, with a portion of the spacecraft that doesn't rotate, perhaps the engine for instance. I've been struggling in my mind to picture how you can put in an interface between the two sections that will allow people to go through without being in a vacuum suit between the two sections, and still be air tight. How would this work in a real operation?

For bonus points, if you could add in how to get power/other cables between the two sections, I would be interested in that as well.

• – TildalWave Oct 7 '13 at 20:07
• If one were to "blue sky" a spacecraft design incorporating artificial gravity via centrifuge, one might consider design options which don't require a dynamic seal at all. One possibility would be to rotate the entire pressure hull, or all of the vehicle except those items which must be stationary e.g. Earth-facing antenna, solar panels, star trackers, etc.. Another option might be to enclose the centrifuge in the pressure vessel, such as the depicted "Discovery" in "2001: A Space Odyssey". – Anthony X Jul 12 '14 at 1:42

The power and data part is easy. The Galileo spacecraft that went to Jupiter was a "dual spinner", which had spun and stationary sections. The rotating joint had 48 slip-rings over which power and data were transmitted between the sections.

As for connecting pressure vessels with a rotating joint, there are similar industrial applications requiring a seal at a rotating joint. I'm sure that something could be worked out with an acceptable leak rate. All spacecraft have such a rate defined. It's never zero.

By the way, it's not clear that you need a despun section for your application. You just need to thrust along your spin axis.

• As far as the last part, I've only heard about spun/despun sections of spacecraft around, I'm not entirely sure how of the whys, but I remember thrusting being a bit tricky, perhaps there were others as well. – PearsonArtPhoto Oct 7 '13 at 20:38
• The ISS outboard truss sections hold the solar arrays and can rotate continuously relative to the main truss...so power is passed back to the main truss via a rolling slip ring type joint. Data passes back and forth in a similar way. I couldn't find a picture in the open literature but the power transfer part of the joint looks like small circles rolling between the inner and outer rings. The ISS thermal radiators, sadly for this question, can't rotate continously. Fluid passes to and from them through hoses that get wound and unwound. – Organic Marble Aug 22 '15 at 2:41

NASA did research on Feasibility of liquid-metal vacuum seals, the article was published in 1963 and is accessible here (please look up section B, page 29).

Here is quote of the articles conclusion sections.

A procedure was found for using a liquid-metal seal in an ultrahigh-vacuum system.

Leak reates through a $$5.5$$ inch-diameter seal were so small that, with a chamber pressure of $$1$$ micron, there was no indication on a leak detector with a sensitivity of $$3*10^{-10}$$ standard cc of helium per second. Also, leakage was small enough to give no appreciable in-bleed to a system at a chamber pressure of $$7*10^{-9}$$ torr.

The liquid-metal seal, essentially eliminating gas permeation and requiring a very small force per unit length to achieve a seal, has several advantages over the conventional O-ring and metallic shear seals. However, because of the extensive preliminary procedure required to prepare the flanges and liquid metal for the seal, the liquid-metal seal would be difficult to make use of on a routine basis.

This technology is considered to be used in Nautilus-X, which as you can see, has a spinning wheel that should provide about a half of Earths gravity experience.

## Edit:

If the portion of the aircraft that is spinning is small enough, the only problems are ones that you listed, power and air pressure, but imagine we are speaking about very big spacecraft with a very big spinning ring (or even more than one big spinning ring), then you would have much more problems, for example liquid exchange (which is usefull for heat exchange or simply water supply for the crew).

• If a spacecraft can dock to the non-spinning part, e.g. through a standard docking port as used on ISS, the tunnel part has an 80 cm diameter. This is fine for people moving through, but should be considered a minimum opening size, much bigger than 5.5 inches. More realistically, you are going to want to move big things in and out of a large habitat, so asking for a 9m+ diameter opening is reasonable. – Terrel Shumway Sep 27 '19 at 22:06

I have an arts degree and am not an engineer, but I work in technical communications and have an aptitude for problem solving, so I'll give it a go.

Marek makes a good point about the acceptable leak rate. I've seen comments from others posted elsewhere that suggest that the seals have to cause friction, or the crew must be in suits to transfer between sections if the seals are loose. That is not necessarily the case.

In the SF tales I'm developing, I accept the limitations of labyrinth seals for preventing leaks of atmosphere and most ships/stations do not normally have the volumes adjacent to the transfer areas pressurized at all. With the air held behind bulkheads and the robust seals of standard hatches, there is no air to leak out through the seals that protect the juncture of the ship sections.

Here are some options for getting crew through the juncture without them getting into suits.

1. Use a pressurized tube with appropriate hatches at the ends that can couple to and detach from the rotating and non-rotating portions of the ship. Crew open one end, enter and close that hatch behind them. The tube detaches and its rotation is changed to match that of the other section, to which it is then connected. The crew open that hatch and exit the transfer tube. No pressure suits or major hassle. Power/data is transferred through slip rings elsewhere, or through cables in the transfer tube wall and slip rings.

2. Instead of just a tube that changes spin, a pod could be transferred through an unpressurized junction and spun to connect with the ship sections as needed, sealing to hatches directly or moving in and out of volumes that can be evacuated and pressurized.

3. A non-rotating tube extends into the rotating section. The adjacent chamber -- perhaps the size of a small room -- is unpressurized when not in use for transferring crew. When that chamber is in use, it is sealed and air is pumped in. Some will leak out, but at a slow rate, and only from that limited volume. When not in use, the transfer tube volume is pumped out to conserve the atmosphere. Again, no air there means no air to leak out through the seals. Some is lost, but at a much slower rate than if it was pressurized at all times. The small cumulative loss of air from the transfer section is balanced with air from the much larger volumes in the rest of the station/vessel. If you really need to pipe water into and out of the rotating section, a pipe or hose could run through an axial transfer tube, down the center in the case of a larger tube or in the wall of the tube so long as it connected to the rotating section at the actual axis with enough room that crew transfer isn't hampered.

4. The tube is optional. Instead, have the two sections meet with two hatches that can be sealed, perhaps a meter apart, or even less. The smaller the volume the faster it can be pumped full or evacuated.

In options 3 and 4, the crew start by pumping air from tanks into the transfer tube section, then open the hatches one at a time or simultaneously, move through into the other section, then close the hatches. If they need to go back and forth, they keep it pressurized, but perhaps closed. If not, they pump the air back into tanks, minus what little amount leaked out in the short time it was in use. Just don't have a phobia about not being able to breathe...

Leakage is not a problem if there is nothing there to leak out.

There is a simple answer to this problem. No seals!

The issues with most spacecraft designs is they are done by aircraft engineers. Ask a ship builder how to make a pressure hull for space, their design will be quite different. All too often do I watch a sci-fi movie and see a space ship with all its key components exposed to the hard environment of space. This looks nice, but both wrong and impractical. If any component failed it would mean sending a person out into open space to fix it. Space walks look like fun but they are extremely risky to do.

So using this line of thinking and the issues of keeping a seal between moving parts in a vacuum. If we designed a spaceship like a super tanker there is no need to put in complex joints that can fail or leak. You simply put the rotating section inside the pressure hull. In fact, you put all the key components inside the hull. Some parts would be unpressurised like the engine space, but all other parts of the cabin and engineering spaces would / could be pressurised (with air or an inert gases) to enable a service engineer to access the area in an environment suit (a more flexible suit that supplies heat, O2). The rotating area within the pressure hull would be in two sections each rotating in the opposing direction, that way the torque effects could be managed by gyroscopes and not by thruster activity.

• There is one basic premise that, um, torpedoes, your idea: Weight matters a lot more in spacecraft than it does in supertankers. – Organic Marble Dec 6 '17 at 13:53
• Mass is less important when you source your material from a shallow gravity well (asteroid). By the time you are ready to build the habitats the OP is talking about, asteroid mining should be a given. Bootstrapping the first mining ship will be a bigger challenge. – Terrel Shumway Sep 27 '19 at 21:58

I was just looking at a clip from The Martian (2015) movie, and their space station looking spacecraft got me thinking about this very thing. I had an idea that I think has been covered in two separate ways but not together, and it doesn't involve any low leak seals to the rotating/stationary section because there wouldn't be any seals. The whole vacuum proof area should rotate, as it has been said, and then everything that needs be stationary would rotate like the solar panels and the heat dissipation radiators and antenna/dish, etcetera could easily rotate outside the vacuum safe section. Two or three rings on the outside of round sections of the hull that counter-rotate as necessary would do the trick. For the electronics, a metal disc rolling along a metal strip. if the engines are in a circular pattern it doesn't matter if they rotate.

• Actually it does matter that engines rotate, because you're losing angular momentum of the fuel on its exhaust products. Even a single engine that's central to the axis of rotation would still introduce torque to the rotating wheel since it wouldn't be a point source of thrust. It wouldn't be too difficult to neutralize that effect tho (cant the nozzles or despin the fuel - LOL I never thought I'm gonna say that), but it's even easier if engines and the fuel aren't rotating at all. – TildalWave Aug 21 '15 at 22:02
• how significant a factor is that angular momentum loss, given that spin stabilised upper stages were a thing for a long time? – JCRM Dec 7 '17 at 11:08
• More good reasons to have a non-spinning part: 1) spinning the shielding adds significant relative velocity to any impacts, requiring more shielding. 2) having two counter-rotating cylinders seems (to me) to eliminate the problem discussed in youtube.com/watch?v=1VPfZ_XzisU. You will at least need some fixed truss to connect the two cylinders. – Terrel Shumway Sep 27 '19 at 22:13