Airlocks for Mars Colony

[MickQ]

As I understand it, naval hatches need to resist pressure from either side whereas an airlock door only has pressure on one side.  The latching mechanism for the airlock therefore could be not so heavy and complicated and use the internal air pressure to provide most of the sealing force once the door is closed.

Is this how things work now ?   If not then will / can it work on Mars and how much pressure would be needed for a good seal ?

[Lampyridae]

If we’re talking about making a gasket in a pinch, ABS plastic (which can be 3D printed and solvent welded together into larger parts and then solvent smoothed to make it airtight) with a little acetone absorbed in it is elastomeric, although it smells awful, would be messy and potentially cause other issues, and wouldn’t last long. But ABS parts have been 3D printed on ISS already and probably we’ll bring this stuff to Mars. (Although I’m not sure anyone has brought a bottle of acetone to ISS… you’d only ever want to use it in a glove box if you’re in a small space habitat for a TON of reasons, although ISS’s life support system can remove acetone vapors from the atmosphere…)

Ferrofluids are used for various rotating seals, with pressures of 30+ KPa, and even for joints extending into vacuum for industrial purposes. You'd want ferrofluid for a variety of mechanical reasons anyway.

So for an airlock -- maybe just have a seal that is magnetisable and it just pulls the ferrofluid across the appropriate groove. Since it's good for rotating joints -- pressure suit rotating joints are super complicated -- maybe it could be used there too. Dust would be a major pain though, and the necessary magnets might be too heavy.

Ironically, ferrofluids are one of the things spun off by NASA and were originally designed for rocket fuel...

[Lampyridae]

I am not sure why the talk of winches with the water sump.

With vehicles you drive down the angled underwater tunnel and then up the other side. Very efficient.

For people on foot you would have a vertical shaft and simply use buoyancy control as used by scuba divers.

Evaporation should not be a problem as a thin layer of oil will prevent this. It might be necessary to have heaters to prevent ice build up when this becomes a problem, but the ice may be useful to prevent water loss when no one is using the passage.

This isn't very practical, but if you have an Archimedes lock (let's call it that) made out of mercury, it could be a lot shorter (also possibly Galinstan, but that forms a skin even in 1ppm O2). A suited person would simply shoot through the liquid like a human cork bottle. Coming down would require a winch or something.

Liquid  1 bar | 1/3 bar
Water: 26m | 8.7m
Mercury: 2m| 0.64m
Galinstan: 4.2m | 1.4m

But since it can be manipulated magnetically, you could just spin the lock open like dialling the Stargate. Pray it doesn't fail and splash you with denser-than-steel liquid, or drag you into the vortex.



Tech level: very old. Literally the first electric motor. Maybe not useful for an actual airlock but perhaps it may help with airlock mechanisms: sealing, pumping etc. Has some nasty interactions with aluminium and oxygen.

To clarify, spinning it Stargate-style is literally just a joke. No need to NSF-quote me and explain why it won't work.

[Twark_Main]

I am not sure why the talk of winches with the water sump.

With vehicles you drive down the angled underwater tunnel and then up the other side. Very efficient.

Not efficient in terms of the quantity of fluid, or in general the overall physical size. Bigger things cost more.

Evaporation should not be a problem as a thin layer of oil will prevent this. It might be necessary to have heaters to prevent ice build up when this becomes a problem, but the ice may be useful to prevent water loss when no one is using the passage.

It should be easy to verify the (in)feasibility of this idea using an inexpensive vacuum pump and some pipe. This is one experiment we can do on Earth, and in a home setting no less!

My expectation: the water will boil near the surface, everywhere above the equipressure surface where the water column pressure drops below the boiling point. It doesn't matter if a thin film means the water can't "see" the vacuum, all that matters is the pressure.  The roiling surface will break up whatever oil or ice layer anyway.

On Mars the goal is to bring precious water inside, not the opposite.  If you thought the water conservation rules implemented in Phoenix were bad, you ain't seen nuthin' yet...

[Lampyridae]

I am not sure why the talk of winches with the water sump.

With vehicles you drive down the angled underwater tunnel and then up the other side. Very efficient.

Not efficient in terms of the quantity of fluid, or in general the overall physical size. Bigger things cost more.

Evaporation should not be a problem as a thin layer of oil will prevent this. It might be necessary to have heaters to prevent ice build up when this becomes a problem, but the ice may be useful to prevent water loss when no one is using the passage.

It should be easy to verify the (in)feasibility of this idea using an inexpensive vacuum pump and some pipe. This is one experiment we can do on Earth, and in a home setting no less!

My expectation: the water will boil near the surface, everywhere above the equipressure surface where the water column pressure drops below the boiling point. It doesn't matter if a thin film means the water can't "see" the vacuum, all that matters is the pressure.  The roiling surface will break up whatever oil or ice layer anyway.

On Mars the goal is to bring precious water inside, not the opposite.  If you thought the water conservation rules implemented in Phoenix were bad, you ain't seen nuthin' yet...

Most oils have a fairly low vapour pressure; most vacuum pump oils should be fine especially as it decreases with temperature anyway. Just pile on sufficient oil to increase the pressure at the top of the water column to above say the Armstrong Limit. That's about 0.6m of Earth water, 1.6m of Mars water so call it 1.8m of oil on Mars. (Vegetable oil seems to boil and emulsifies with water, great for making salad dressings I guess)

PFPE "oils" are used in vacuum pumps, spacecraft etc. Only trouble is their 1.8g/cc density so you can't float it over some water. But you could make an Archimedes lock of water connect to a secondary lock (entrance above the Armstrong limit) of PFPE. PFPE can also be a ferrofluid carrier.

Ionic liquids can also have very low vapour pressures. Some can be made to be paramagnetic. They have been proposed for liquid mirror telescopes on the Moon¹ – one study² coated an ionic liquid with silver and found that it was stable for months. NASA is actively researching liquid mirror telescopes.³

I don't know if a magnetic fluid could be manipulated to press against atmospheric pressure. I know this effectively happens in a manner with the aforementioned vacuum seal, but whether a "liquid door" would be possible by electromagnetically restraining a magnetic fluid is possible, I don't know. Ferrofluids have some tricky behaviours and they're not really used in bulk.

1. https://phys.org/news/2008-10-liquid-mirror-telescopes-moon.html
2. https://www.nature.com/articles/nature05909
3. https://www.nasa.gov/science-research/astrophysics/what-is-the-fluidic-telescope/

[Twark_Main]

Most oils have a fairly low vapour pressure; most vacuum pump oils should be fine especially as it decreases with temperature anyway. Just pile on sufficient oil to increase the pressure at the top of the water column to above say the Armstrong Limit. That's about 0.6m of Earth water, 1.6m of Mars water so call it 1.8m of oil on Mars. (Vegetable oil seems to boil and emulsifies with water, great for making salad dressings I guess)

PFPE "oils" are used in vacuum pumps, spacecraft etc. Only trouble is their 1.8g/cc density so you can't float it over some water. But you could make an Archimedes lock of water connect to a secondary lock (entrance above the Armstrong limit) of PFPE. PFPE can also be a ferrofluid carrier.

Why even bother with the water? Just have it be turtles PFPE oils the whole way down.

This also reduces the height of your shaft or ramp by 1 - 1.0/1.8 = 44%. If the large construction is only half the size (actually even better, because you can eliminate the Archimedes lock), it could be worth it to use more expensive oil instead of water.

[Paul451]

You don't need oil on the water to reduce evaporation/boiling, you'd have an unpressurised outer chamber that is saturated with water vapour. An outer door (not hatch) reduces mixing with the outer atmosphere. (For vehicle size water-locks, you might get away with air-curtain type barriers, avoiding the need for an outer door at all.)

[Twark_Main]

You don't need oil on the water to reduce evaporation/boiling, you'd have an unpressurised outer chamber that is saturated with water vapour.

That won't do much to slow evaporation. The water will continue boiling and build up pressure inside the chamber. You just built a steam engine. 

[Lampyridae]

Why even bother with the water? Just have it be turtles PFPE oils the whole way down.

This also reduces the height of your shaft or ramp by 1 - 1.0/1.8 = 44%. If the large construction is only half the size (actually even better, because you can eliminate the Archimedes lock), it could be worth it to use more expensive oil instead of water.

Expediency. You don't need to have 100 or so tonnes of PFPE/ionic fluids etc for a human-sized airlock, you just need some water. Plus suits and humans are quite buoyant in higher densities as you pointed out, so the inbound traverse is difficult.

You don't need oil on the water to reduce evaporation/boiling, you'd have an unpressurised outer chamber that is saturated with water vapour. An outer door (not hatch) reduces mixing with the outer atmosphere. (For vehicle size water-locks, you might get away with air-curtain type barriers, avoiding the need for an outer door at all.)

Perhaps just use something like the simple fabric airlocks on inflatable structures on Earth, just inflated with local atmosphere. The pressure differential normally used with those things is like 2-3kPa which is not fantastic; not enough to get even to 0°C boiling limit above Mars pressure.

[Paul451]

Perhaps just use something like the simple fabric airlocks on inflatable structures on Earth, just inflated with local atmosphere. The pressure differential normally used with those things is like 2-3kPa which is not fantastic; not enough to get even to 0°C boiling limit above Mars pressure.

Am I reading water's phase diagram wrong?

The boiling point of water at 2 kPa is ~14°C. At 3 kPa it is ~20°C (or ~24° when including Mars ambient.)

[Aside: My only issue with the outer lock partially-pressurised with Mars atmosphere is that you really want that outer lock at saturation humidity to reduce evaporation. A low-pressure inflatable pumping Mars air assumes constant air loss, which means humidity will be low, which means constant water loss. Not boiling, just normal dry-air evaporation. But I was hoping to avoid any significant water loss. You could still have the actual outer lock in a room/garage at high humidity, then an unpressurised door into your inflatable Mars-air tent to reduce mixing. Still mechanically simpler than a normal airlock, and still much higher traffic throughput (which was the point.)]


Speaking of reading the phase diagram wrong, I just googled for an online phase calculator for water, to double check myself, and, of course, google's AI helpfully worked it out for me... yay! ...telling me the 3 kPa boiling point was 69.1° C.

So there!

AI saves the day!


Out of curiosity, I went back and tried the AI at 2 kPa, and this time it quoted the Clausius-Clapeyron equation, and nearly got the right temperature, 14.1°C instead of 13.99°C. Minor rounding error? So I tried 1 kPa, and this time it even showed the Clausius-Clapeyron equation! Correctly! And then showed the steps to use it! Buuuut it got the answer wrong, 6.35°C instead of 2.7°C, much more than a rounding error. So I tried 3 kPa again... still 69.1°C, but this time I realised that 69.1° Fahrenheit is awfully close to the right answer in Celsius. Tried 4 kPa, got 28.96°C instead of 26°C. Then I tried 5 kPa, and got

$32.87^\circ$C...

So now I'm closing the tab before I hurt something.

imtiredboss.gif

[Twark_Main]

Speaking of reading the phase diagram wrong, I just googled for an online phase calculator for water, to double check myself, and, of course, google's AI helpfully worked it out for me... yay! ...telling me the 3 kPa boiling point was 69.1° C.

So there!

AI saves the day!

If you want the right answer, Wolfram|Alpha can do that.

https://www.wolframalpha.com/input?i=boiling+point+of+water+at+3+kPa

[Paul451]

If you want the right answer, Wolfram|Alpha can do that.
https://www.wolframalpha.com/input?i=boiling+point+of+water+at+3+kPa

Weird, that gets it wrong too.

[Edit: I mean, assuming we're using R = 8.315 J/mol*K, ΔH_vap for water = 40.657 kJ/mol, and a baseline of 100°C at 101.325kPa, and the standard Clausius-Clapeyron equation.]

[Twark_Main]

I just googled for an online phase calculator for water, to double check myself

If you want the right answer, Wolfram|Alpha can do that.

https://www.wolframalpha.com/input?i=boiling+point+of+water+at+3+kPa

Weird, that gets it wrong too.

[Edit: I mean, assuming we're using R = 8.315 J/mol*K, H_vap for water = 40.657 kJ/mol, and a baseline of 100°C at 101.325kPa, and the standard Clausius-Clapeyron equation.]

I expect it's using a more exact (non-ideal) model of water.

You're trying to "double check yourself," but if you're assuming your answer is right and anything else is wrong then... what are you double-checking exactly? 

Is it just that you doubt your own arithmetic?  I use an exact unit-preserving calculator to avoid this, but presumably some folks are still stuck on Calc.exe or Excel or similar where you're expected to "raw dog" the units all by yourself.  That always seems totally bonkers to me,

[Paul451]

If you want the right answer, Wolfram|Alpha can do that.

Weird, that gets it wrong too.
[Edit: I mean, assuming we're using R = 8.315 J/mol*K, ΔH_vap for water = 40.657 kJ/mol, and a baseline of 100°C at 101.325kPa, and the standard Clausius-Clapeyron equation.]

I expect it's using a more exact (non-ideal) model of water.
You're trying to "double check yourself," but if you're assuming your answer is right and anything else is wrong then... what are you double-checking exactly?

I was double checking my eyeballing a phase change chart against a calculated result by an online tool. Then I checked those results by hand using the CC equation, noting that most online tools get the same result that I do, but Wolfram-Alpha doesn't (nor does Google AI, by a much wilder and more amusingly random margin.)

Is it just that you doubt your own arithmetic?

No, I doubted the claim by Lampyridae that 3 kPa over Mars ambient isn't enough to push the boiling point of water above 0°C. Then I doubted my ability to draw two converging lines on a chart. Then I doubted the online tools. Then I doubted Wolfram Alpha.

And it turns out my arithmetic was fine. But...

assuming we're using [...] ΔH_vap for water = 40.657 kJ/mol

No, we are not. It varies with temperature. That's annoying. It's better to use ~44 kJ/mol for temps between 0 and 30°C. (Used a different online tool. Good old Engineering Toolbox. Dumb as a box of hammers, but also as reliable.)

The whole thing makes using any equation that uses ΔH_vap pretty useless for calculating boiling points, since you need to calculate ΔH_vap for the temperature that you are trying to find. Very circular. Curious what the best equation would be. Can't figure out a question-phrasing that gets WA to tell me what equation it's using. (And google is still worthless.)

[Also noted that while WA quotes the same ΔH_vap figure for near-100°C as everyone else (although it struggles with converting to J/mol), it must be hard coded (or hard data'd?) to use a different figure when actually calculating boiling point. But I can't get it to show what. That hard coded data might be right, but I have no way to check its work.]

[Robotbeat]

If you want the right answer, Wolfram|Alpha can do that.

Weird, that gets it wrong too.
[Edit: I mean, assuming we're using R = 8.315 J/mol*K, ΔH_vap for water = 40.657 kJ/mol, and a baseline of 100°C at 101.325kPa, and the standard Clausius-Clapeyron equation.]

I expect it's using a more exact (non-ideal) model of water.
You're trying to "double check yourself," but if you're assuming your answer is right and anything else is wrong then... what are you double-checking exactly?

I was double checking my eyeballing a phase change chart against a calculated result by an online tool. Then I checked those results by hand using the CC equation, noting that most online tools get the same result that I do, but Wolfram-Alpha doesn't (nor does Google AI, by a much wilder and more amusingly random margin.)

Is it just that you doubt your own arithmetic?

No, I doubted the claim by Lampyridae that 3 kPa over Mars ambient isn't enough to push the boiling point of water above 0°C. Then I doubted my ability to draw two converging lines on a chart. Then I doubted the online tools. Then I doubted Wolfram Alpha.

And it turns out my arithmetic was fine. But...

assuming we're using [...] ΔH_vap for water = 40.657 kJ/mol

No, we are not. It varies with temperature. That's annoying. It's better to use ~44 kJ/mol for temps between 0 and 30°C. (Used a different online tool. Good old Engineering Toolbox. Dumb as a box of hammers, but also as reliable.)

The whole thing makes using any equation that uses ΔH_vap pretty useless for calculating boiling points, since you need to calculate ΔH_vap for the temperature that you are trying to find. Very circular. Curious what the best equation would be. Can't figure out a question-phrasing that gets WA to tell me what equation it's using. (And google is still worthless.)

[Also noted that while WA quotes the same ΔH_vap figure for near-100°C as everyone else (although it struggles with converting to J/mol), it must be hard coded (or hard data'd?) to use a different figure when actually calculating boiling point. But I can't get it to show what. That hard coded data might be right, but I have no way to check its work.]

Wolfram Alpha uses NIST data as its basis for stuff like this. You can also check with NIST directly.

[Twark_Main]

Any progress with someone actually testing this?

As far as I can tell all you'd need a 35 foot tall (ideally clear) S-shaped tube with a U-bend at the top and bottom. Prop up vertically. Fill the lower U-bend side with 18' of water and top with your preferred vacuum-rated oil (using a smaller S-bend if needed), this is the "airlock" side. Cap the the other leg and adapt it to ~500 pascal vacuum pump and gauge, this is the "Martian atmosphere" side. Purge with CO2 ff desired for higher fidelity.

Some Mars demonstrators require a large effort, but literally someone with a decent garage could advance this one to TRL3 in a weekend.