Elon Musk: glass geodesic domes

[lamontagne]

Hi everybody. I resurrect this old subject. There is something people didn't think about in the 28 previous pages : the ductile/fragile transition for the steel. Medium temperature on Mars is -60 ° C, just the temperature to which steel becomes fragile -> it breaks.
So, what can we do ? Warm the steel thanks to little warming chords in it ?

Furthermore, another issue : the wind. It can destroy the dome with natural frequency.

Another thing : https://en.wikipedia.org/wiki/Creep_(deformation)   The creep. On a long time (20 years), steel will strongly deformate, so the lifetime of the dome will be limited, it is sad.

Hello Poly, I see it is your first post, so welcome to the forum!

-As far as fragile/ductile, there are different qualities of steel, for this example the transition temperature starts at -90C, so it would be OK : https://www.sciencedirect.com/topics/engineering/brittle-transition-temperature
Which raises the interesting question, what is the steel required, and in particular the alloys added to get this lower transition temperature steel and how common are these materials on Mars.

-Yes, like the Tacoma  narrows bridge?  There is very little energy in the Martian winds, since the density of the air is so low.  but it would be a good idea to at least look at the equations and see if they can be applied.
-Creep is more of a high temperature phenomenon.  As the metal gets colder, it creeps less. Do you think creep would be significant enough to distort the structure and cause problems?  the glass in the dome needs to be on some kind of polymer/flexible support anyway. 
Somewhere in this thread is the design used for the lunar LEM.

-There is also the issue of metal fatigue.  The day night cycle of Mars is a cyclical loading and therefore should lead to metal fatigue.  That is one of the good point of steel, since it has a long life cycle and even reaches a plateau.  Aluminum, for example, is more prone to fatigue.

In a dome, the interior temperature will be 20°C or more, so the metal forming the dome will be somewhere between 20°C and -68°C.  The actual temperature of the steel depends on the insulation between it and the interior air, and the radiative temperature of the surface of the steel when you take the convective and radiative effects into account.  Steel is a poor insulator, so the surface of the steel should be quite a bit higher than -68°C.

So basically, to solve the questions you need to answer: what allow would have a lower transition temperature, how much added allowance must we make for creep and if the energy in the wind sufficient to induce waves in the dome.

[lamontagne]

I suggested glass safety film a few pages ago. Surprisingly thin, just a few mills, and it is very tough.  Simply stuck to the glass.  It could come from Earth at first, then be produced locally.

The adhesive layer will possibly present some problems over time. the application is way off spec for it, is the obvious suspicion. That adhesive is specified for human earth application.

Where... beginning an aging test for use on mars, is something that should be considered now, not later. Accelerated aging - testing can only go so far, with regard to potential revelations. Have a friend who worked in BASF coatings age testing for 17 years, just some of his off-cuff casually conferred learnings. And the adhesives involved in such technologies...might likely come from BASF.

That the adhesive and the material itself might ahve to be re-specified for Mars conditions, and those conditions make a customized adhesive used on earth, well, those conditions might make short shrift of the idea of simple technological utility transfer. 

Whereas if you re-specify and re-design.. you head right into untested chemical combinations that might not be half as serviceable on mars as they might be on earth.

Unknowns, in a pressurized dome, over your head, on Mars, does not seem like the best idea. Thus, design and test now, is the minimum required - to go forward with such thinking.

The only gift given, says that the stabilizing layer probably goes to the inside surface, not the outside, so some conditions of exposure are mitigated to some degree.

Absolutely.  Even the UV conditions are different.  to be honest I'm puzzled SpaceX hasn't already built a large scale testing chamber/facility to check out some of these things, in particular those that take time.  the giant Starship window would be an obvious design that would need to be tested fairly soon.  Without testing all these speculations can just go so far, and then hang about, unresolved....

[spacenut]

If steel gets fragile enough to break, how do they get the cold liquid oxygen and liquid methane to not destroy their stainless steel rocket? 

Many smaller double or triple paned glass panels could be installed in a greenhouse dome.  If something breaks one, it should hold until a replacement is installed.  Also, the glass could be made similar to auto windshield glass, layered and glued together.  Even a small wire grid could be sandwiched between the glass for more strength and security from breakage. 

I would think etching by sand and wind over time would cause the glass to be hazed enough to either be replaced or polished out. 

A retractable canopy could cover the greenhouses at night for extra protection from wind and meteorites.  However, some people may want to see the sky at night from inside a greenhouse dome. 

Oh, a broken glass could be removed and replaced outside like they do for cracked car windshields.  Same could also be done from the inside with a double pane system.  Each panel could also have a place to pull a vacuum between the glass panes for thermos type insulation of the greenhouse. 

The protective night cover could be as simple as a large polished piece of stainless steel to reflect more light into the greenhouse in the day while covering it at night.  Greenhouses could be made 9m in diameter to reused expended parts of Starships for construction. 

[Oersted]

to be honest I'm puzzled SpaceX hasn't already built a large scale testing chamber/facility to check out some of these things, in particular those that take time.  the giant Starship window would be an obvious design that would need to be tested fairly soon. 

First things first. In any case, those big acrylic windows will go the same way as the helicopter-like windows on the early Lunar LEM designs. Too heavy and fragile and in any case you get just as wide a field of view by pressing your nose up to a small window.

I'd say that the same thing goes for big domes on Mars, but we've discussed that to bits over 150+ pages of "envisioning amazing habitats". 

[Poly]

If steel gets fragile enough to break, how do they get the cold liquid oxygen and liquid methane to not destroy their stainless steel rocket? 

Many smaller double or triple paned glass panels could be installed in a greenhouse dome.  If something breaks one, it should hold until a replacement is installed.  Also, the glass could be made similar to auto windshield glass, layered and glued together.  Even a small wire grid could be sandwiched between the glass for more strength and security from breakage. 

I would think etching by sand and wind over time would cause the glass to be hazed enough to either be replaced or polished out. 

A retractable canopy could cover the greenhouses at night for extra protection from wind and meteorites.  However, some people may want to see the sky at night from inside a greenhouse dome. 

Oh, a broken glass could be removed and replaced outside like they do for cracked car windshields.  Same could also be done from the inside with a double pane system.  Each panel could also have a place to pull a vacuum between the glass panes for thermos type insulation of the greenhouse. 

The protective night cover could be as simple as a large polished piece of stainless steel to reflect more light into the greenhouse in the day while covering it at night.  Greenhouses could be made 9m in diameter to reused expended parts of Starships for construction.

A retractable canopy is expensive, so, better not do that. We can send a forge on Mars for 40 T of materials, but only to make bars and little elements, not giant plates of steel.

Furthermore, I thought about something. The link steel-glass on a geodesic dome on mars is very simple. As I drew it in the picture, with some kind of caoutchouc-like materials (joint), if glass thermically contracts or expands, it shrinks (or expands) the caoutchouc. Pressure maintains the glass panel. Do you see any issue with this ?

And about polymers, no, we must not use polymers. Polymers turn yellow because of UV and are easily scratched by the sand, contrary to though glass.

[Paul451]

Things to consider:

1) You don't need as much support under the glass as over it. The largest force is outwards. The only support you need under the glass is that required to support the weight of the glass when the dome is unpressurised. While pressurised, the glass is transferring 10 tonnes per square metre of glass area to the front of the frame. Similarly, the seal needs to be primarily between the outer face of the glass and the front of the frame. You won't have a symmetrical frame.

However, that might simplify dealing with the differences in expansion between the glass and the steel frame. The glass can, in essence, "float" against its outer seal, pressed against the seal by the outward pressure, with only small catches to support it while unpressurised. Hence as temperatures change, the glass can slide back and forth against its seal, without the seal having to absorb that expansion.

2) Similarly, the frame is primarily tensile. That reduces the advantage of traditional bending-resistant structures like I-beams, used for compressive loads. In addition, to reduce heat loss through the frame, you would want a structure that has thermal barriers within the frame. (Ditto the glass, see Number 4, below.)

3) You need to be able to get the glass in and out for maintenance. As shown, you would have to dismantle the whole framework to replace a single pane.

4) You would want multiple layers of panes. By "layers", I don't just mean laminated panes, but multiple separate air-gapped panes. This gives you improved insulation, reduces the risk of a catastrophic break, makes maintenance easier, and allows you to optimise the panes for their role. For example, the innermost is exposed to moist oxygenated air plus microflora, outermost is exposed to near vacuum, chemically and mechanically corrosive dust, UV, temperature extremes, etc. There are likely to be different optimised coatings/etc.

[Poly]

Things to consider:

1) You don't need as much support under the glass as over it. The largest force is outwards. The only support you need under the glass is that required to support the weight of the glass when the dome is unpressurised. While pressurised, the glass is transferring 10 tonnes per square metre of glass area to the front of the frame. Similarly, the seal needs to be primarily between the outer face of the glass and the front of the frame. You won't have a symmetrical frame.

However, that might simplify dealing with the differences in expansion between the glass and the steel frame. The glass can, in essence, "float" against its outer seal, pressed against the seal by the outward pressure, with only small catches to support it while unpressurised. Hence as temperatures change, the glass can slide back and forth against its seal, without the seal having to absorb that expansion.

2) Similarly, the frame is primarily tensile. That reduces the advantage of traditional bending-resistant structures like I-beams, used for compressive loads. In addition, to reduce heat loss through the frame, you would want a structure that has thermal barriers within the frame. (Ditto the glass, see Number 4, below.)

3) You need to be able to get the glass in and out for maintenance. As shown, you would have to dismantle the whole framework to replace a single pane.

4) You would want multiple layers of panes. By "layers", I don't just mean laminated panes, but multiple separate air-gapped panes. This gives you improved insulation, reduces the risk of a catastrophic break, makes maintenance easier, and allows you to optimise the panes for their role. For example, the innermost is exposed to moist oxygenated air plus microflora, outermost is exposed to near vacuum, chemically and mechanically corrosive dust, UV, temperature extremes, etc. There are likely to be different optimised coatings/etc.

1) You're right.

3) I 've modified the drawing.

4) A break would not be a catastrophy would it ? Air would not flee away with a big BOOM through a tiny hole in the dome. It would be slow, as when a tire has a hole, so that we have some time to replace pannels. But you're right about the optimisation of the panels. I will think about it. Is the optimisation worth the cost of building another layer ? I have to calculate.

[Arb]

Relevant: Crew replace pane on ISS Cupola window written by Chris Bergin April 15, 2017

From which the below image:

[lamontagne]

Things to consider:

1) You don't need as much support under the glass as over it. The largest force is outwards. The only support you need under the glass is that required to support the weight of the glass when the dome is unpressurised. While pressurised, the glass is transferring 10 tonnes per square metre of glass area to the front of the frame. Similarly, the seal needs to be primarily between the outer face of the glass and the front of the frame. You won't have a symmetrical frame.

However, that might simplify dealing with the differences in expansion between the glass and the steel frame. The glass can, in essence, "float" against its outer seal, pressed against the seal by the outward pressure, with only small catches to support it while unpressurised. Hence as temperatures change, the glass can slide back and forth against its seal, without the seal having to absorb that expansion.

2) Similarly, the frame is primarily tensile. That reduces the advantage of traditional bending-resistant structures like I-beams, used for compressive loads. In addition, to reduce heat loss through the frame, you would want a structure that has thermal barriers within the frame. (Ditto the glass, see Number 4, below.)

3) You need to be able to get the glass in and out for maintenance. As shown, you would have to dismantle the whole framework to replace a single pane.

4) You would want multiple layers of panes. By "layers", I don't just mean laminated panes, but multiple separate air-gapped panes. This gives you improved insulation, reduces the risk of a catastrophic break, makes maintenance easier, and allows you to optimise the panes for their role. For example, the innermost is exposed to moist oxygenated air plus microflora, outermost is exposed to near vacuum, chemically and mechanically corrosive dust, UV, temperature extremes, etc. There are likely to be different optimised coatings/etc.

1) You're right.

3) I 've modified the drawing.

4) A break would not be a catastrophy would it ? Air would not flee away with a big BOOM through a tiny hole in the dome. It would be slow, as when a tire has a hole, so that we have some time to replace pannels. But you're right about the optimisation of the panels. I will think about it. Is the optimisation worth the cost of building another layer ? I have to calculate.

There is an equation for the speed of the pressure drop, if you want I have a version somewhere not too far.  It would be very noisy, but it will take quite a bit of time to empty a dome.
Many layers are also interesting for temperature control.  You can use the vacuum as insulation and reduce radiation losses using infra red reflective coatings.  The glass becomes a kind of multi layer insulation.

[Poly]

Things to consider:

1) You don't need as much support under the glass as over it. The largest force is outwards. The only support you need under the glass is that required to support the weight of the glass when the dome is unpressurised. While pressurised, the glass is transferring 10 tonnes per square metre of glass area to the front of the frame. Similarly, the seal needs to be primarily between the outer face of the glass and the front of the frame. You won't have a symmetrical frame.

However, that might simplify dealing with the differences in expansion between the glass and the steel frame. The glass can, in essence, "float" against its outer seal, pressed against the seal by the outward pressure, with only small catches to support it while unpressurised. Hence as temperatures change, the glass can slide back and forth against its seal, without the seal having to absorb that expansion.

2) Similarly, the frame is primarily tensile. That reduces the advantage of traditional bending-resistant structures like I-beams, used for compressive loads. In addition, to reduce heat loss through the frame, you would want a structure that has thermal barriers within the frame. (Ditto the glass, see Number 4, below.)

3) You need to be able to get the glass in and out for maintenance. As shown, you would have to dismantle the whole framework to replace a single pane.

4) You would want multiple layers of panes. By "layers", I don't just mean laminated panes, but multiple separate air-gapped panes. This gives you improved insulation, reduces the risk of a catastrophic break, makes maintenance easier, and allows you to optimise the panes for their role. For example, the innermost is exposed to moist oxygenated air plus microflora, outermost is exposed to near vacuum, chemically and mechanically corrosive dust, UV, temperature extremes, etc. There are likely to be different optimised coatings/etc.

1) You're right.

3) I 've modified the drawing.

4) A break would not be a catastrophy would it ? Air would not flee away with a big BOOM through a tiny hole in the dome. It would be slow, as when a tire has a hole, so that we have some time to replace pannels. But you're right about the optimisation of the panels. I will think about it. Is the optimisation worth the cost of building another layer ? I have to calculate.

There is an equation for the speed of the pressure drop, if you want I have a version somewhere not too far.  It would be very noisy, but it will take quite a bit of time to empty a dome.
Many layers are also interesting for temperature control.  You can use the vacuum as insulation and reduce radiation losses using infra red reflective coatings.  The glass becomes a kind of multi layer insulation.

Yes, I would be glad to get this equation please.

[Vanspace]

Many layers are also interesting for temperature control.  You can use the vacuum as insulation and reduce radiation losses using infra red reflective coatings.  The glass becomes a kind of multi layer insulation.

On earth inert gas is sometimes used for high end windows instead of vacuum between panes for temperature control. Since ISRU will have to process Mars atmosphere anyway, there is an available source for Argon. Could that be used in this application?

[Paul451]

On earth inert gas is sometimes used for high end windows instead of vacuum between panes for temperature control. Since ISRU will have to process Mars atmosphere anyway, there is an available source for Argon. Could that be used in this application?

CO₂ will achieve the same thing. Lets visible light through but absorbs IR trying to radiate back out.

If you have layers, put internal-pressure CO₂ between the inner layers, then low pressure (Mars ambient) CO₂ between the outer layers. The middle layer(s) being the main structural layer. (Letting you optimise the properties of each layer.)

[Paul451]

4) A break would not be a catastrophy would it ? Air would not flee away with a big BOOM through a tiny hole in the dome. It would be slow

If the glass fails, the whole pane would probably fail. Even if each pane is laminated with polymer to bind it, the loss of structural strength from the shattered glass would probably tear it out of the frame.

So how fast it leaks is the average velocity of the air molecules times the air pressure (or density) times the cross sectional area of the broken pane. The mean velocity of air molecules at room temperature is a bit above supersonic (about 500m/s), but you can use the speed of sound at that temperature (300-350m/s) as a rough approximation for annoying statistical reasons. As the pressure drops, both the density and temperature are dropping, but it's not significant enough to change the result by more than an order of magnitude, so you can just use the overall volume and divide by the initial rate of leakage to get a rough approximation. However, I don't have a neatly pre-calculated formula to use.

[edit: Oh wait, yes I do. For this simplification, it's just the volume divided by the area of the hole, then divided by the molecular velocity.
Ie, Time = Volume / (Area X Velocity), being sure to match your units. Cubic and square metres, seconds, and metres/second.
It is only accurate for the bulk of the leak, not the final depressurisation to ambient pressure. But we only care about the first 90% of air loss, so it's good enough. Then halve it to give you "time until you die". Halve it again for safe evacuation.
Example: 500m wide hemispherical dome, 1 square metre broken pane. 18 hours to total air loss, call it 9hrs 'til death. Say 4-5hrs to evacuate the dome and/or repair the break.]

[Annoying statistical reasons: The distribution curve is hard capped on one side at zero, but has no upper limit, giving it a bias. So both the mean and RMS are higher than the peak of the curve. The "most probable" speed is the speed of sound, but the distribution is lop-sided.]

[Poly]

4) A break would not be a catastrophy would it ? Air would not flee away with a big BOOM through a tiny hole in the dome. It would be slow

If the glass fails, the whole pane would probably fail. Even if each pane is laminated with polymer to bind it, the loss of structural strength from the shattered glass would probably tear it out of the frame.

So how fast it leaks is the average velocity of the air molecules times the air pressure (or density) times the cross sectional area of the broken pane. The mean velocity of air molecules at room temperature is a bit above supersonic (about 500m/s), but you can use the speed of sound at that temperature (300-350m/s) as a rough approximation for annoying statistical reasons. As the pressure drops, both the density and temperature are dropping, but it's not significant enough to change the result by more than an order of magnitude, so you can just use the overall volume and divide by the initial rate of leakage to get a rough approximation. However, I don't have a neatly pre-calculated formula to use.

[edit: Oh wait, yes I do. For this simplification, it's just the volume divided by the area of the hole, then divided by the molecular velocity.
Ie, Time = Volume / (Area X Velocity), being sure to match your units. Cubic and square metres, seconds, and metres/second.
It is only accurate for the bulk of the leak, not the final depressurisation to ambient pressure. But we only care about the first 90% of air loss, so it's good enough. Then halve it to give you "time until you die". Halve it again for safe evacuation.
Example: 500m wide hemispherical dome, 1 square metre broken pane. 18 hours to total air loss, call it 9hrs 'til death. Say 4-5hrs to evacuate the dome and/or repair the break.]

[Annoying statistical reasons: The distribution curve is hard capped on one side at zero, but has no upper limit, giving it a bias. So both the mean and RMS are higher than the peak of the curve. The "most probable" speed is the speed of sound, but the distribution is lop-sided.]

Ok, thank you ! I calculated for smaller domes, and air-leak takes hours. If the services of the colony are not too bureaucratic, it should be okay even if the dome is a single-layered one.

I have another issue. Material fatigue can take place in extension, can't it ?

[lamontagne]

Yes, but it's rarely a limit case.  You should proably plan for a maximum lifetime of X, and plan the recycling of the dome from the start.

[Robotbeat]

I think glass domes don't make sense really. polymer much more forgiving. Even so, you could engineer panes to handle a hole without breaking. Just make it double layer with polymer in the middle like safety glass. Heck, adding a thick polymer layer provides some radiation shielding as well. The polymer could be the primary load bearer, with the glass only occurring to maintain optical clarity (plastic isn't as resistant to scratching and sand abrasion).

I tend to think big ETFE bubbles will be the solution, though.


there is a strong case for a decent sized, high-optical-clarity dome on mars, tho. think like a huge Cupola. small windows on ISS just do not compare to the glory of the Cupola. Martians will want to look at the stars. And Earth!

forgive the distorted screenshots. i am in vr.

[lamontagne]

Pressure drop approximate equation.

[Paul451]

I think glass domes don't make sense really.

Well, domes don't make any sense. Doesn't matter what you make them out of.

polymer much more forgiving.

That's a problem. They distort over time. Even nominally "non-stretch" polymers experience creep when subject to continuous tensile loads, which not only distorts the material, it reduces its strength.

[Poly]

I think glass domes don't make sense really.

Well, domes don't make any sense. Doesn't matter what you make them out of.

polymer much more forgiving.

That's a problem. They distort over time. Even nominally "non-stretch" polymers experience creep when subject to continuous tensile loads, which not only distorts the material, it reduces its strength.

Domes makes no sense for robots. Humans who will pay to live on Mars will want to live in a sci-fi movie, with DOMES. The important point is to sell dreams to people.

[Paul451]

I think glass domes don't make sense really.

Well, domes don't make any sense.

Domes makes no sense for robots. Humans who will pay to live on Mars will want to live in a sci-fi movie, with DOMES.

I meant from a structural point of view. Domes are a bad structure to use as a pressure vessel.

The important point is to sell dreams to people.

It's important not to deceive them, however.