Speculation: Segmented Spin gravity habitat sized for launch in 2017 BFR Cargos

[rakaydos]

So the BFR will offer a whole new level of payload to orbit. Is it enough to build a pre-assembled Oneill Colony? Spin gravity and all?


Assuming the framework can be made light enough that lift mass isnt an issue... By eyeball, the BFR Cargo's potetial cargo area is  a cone just under 9m wide (that skin is ridiculusly thin) and possily as much as 1/3 the ship's total 48m length, or 16m long... but that's a cone, and you need pie pieces to build a cylender. So cut the length short to fit 4m square at the narrow end, and 4m by 8m at the wide end... Aim for 8 segments per ring, 1 segment per launch.

At those dimentions, with 8 segments, the hollow donuthole is just over 9.5m at the narrowest according to my calculations (octogon with 4m sides, 4 meters + 2 x(2 meters x sqrt(2) )= 9.657 meters) allowing a BFR to dock at the center of the ring, 2001 Space Oddessy style. The outside of the ring is twice that at 18m, which can generate a mars gravity (.3g) at just under 4 RPM.

Is this concept more than just a pie on the sky?

[QuantumG]

http://www.nss.org/settlement/space/GlobusEasierSettlement.pdf

Worth reading.

[docmordrid]

I would think that this would be a project for Teathers Unlimited's SpiderFab.

https://www.nasa.gov/directorates/spacetech/niac/2012_phase_I_fellows_hoyt_spiderfab.html

http://www.tethers.com/SpiderFab.html

[KelvinZero]

So you are talking about a toroid, not one of those big cylinder things? What constraints are you aiming for? 2rpm or better? Full earth gravity or only Mars gravity?

(Elon has discussed a much more far fetched proposal: massive geodesic domes on mars made from components transported in BFS cargo. I vaguely recall glass panes. Anyway, if you can do an arbitrary sized dome then you can do an arbitrary sized sphere for example.)

We could just ask what you could assemble with Bigelow components just as examples, just to constrain it further.

[rakaydos]

http://www.nss.org/settlement/space/GlobusEasierSettlement.pdf

Worth reading.

Indeed, though it seems focused on 1g long term habitats, which I suppose was impled by my "oneill cylender" comment.

The design I put forth, though, is an ISS-level space habitat intended for up to 1/3g for human physioligy testing, with a discreet number of segmets constructed before launch and docked together in orbit before spinning up.

Spiderfab looks interesting, but I hope to invalidate it with pre-assembily.

[rakaydos]

So you are talking about a toroid, not one of those big cylinder things? What constraints are you aiming for? 2rpm or better? Full earth gravity or only Mars gravity?

(Elon has discussed a much more far fetched proposal: massive geodesic domes on mars made from components transported in BFS cargo. I vaguely recall glass panes. Anyway, if you can do an arbitrary sized dome then you can do an arbitrary sized sphere for example.)

We could just ask what you could assemble with Bigelow components just as examples, just to constrain it further.

The contraints I'm looking at is how shallow a pie-slice will fit in the BFR cargo. The shallower the angle, the more segments per ring and so the bigger the assemled ring is when it's spun up, allowing higher g/lower RPM, as well as more habitable area.

[speedevil]

The cylindrical part of BFS, around the airlock can carry a cylinder 9m diameter, by 8m long.
It can lift around 150 tons.
This is 75m^3 of aluminium or so.
If we assume flat 8*8m plates with minimum curvature, that is 1*8*8m or so overall.
For no particularly good reason, imagine they are 25mm thick.
We can carry 40 of them.

Assume working load is 150MPa, and that internal pressure is the same as Flagstaff Az (75KPa), that means that the workable diameter is 100m.

Let us ignore that for the moment, and assume we want to make a section of cylinder about the same diameter it is long.

For 40 panels, you can make a cylinder 30m in diameter, and 24m long.

The structural loading on this is about 15% of yield, meaning the joining can have quite large stress concentrations without issue.
Ten launches get you the shell of a large toroidal station with the outside diameter 80m, with a cylinder diameter of 30m.
Or a cylinder 200m or so long by 30.

It would need a whole full load of liquid air to fill it to STP, around 150 tons.

You then begin to outfit the inside, once in STP.

(toroidal station of course would not have the segments quite square, and they would be individual.)

Joins would be handled with two million five hundred thousand M8 bolts, all alone in the night.

[rakaydos]

The cylindrical part of BFS, around the airlock can carry a cylinder 9m diameter, by 8m long.
It can lift around 150 tons.
This is 75m^3 of aluminium or so.
If we assume flat 8*8m plates with minimum curvature, that is 1*8*8m or so overall.
For no particularly good reason, imagine they are 25mm thick.
We can carry 40 of them.

Assume working load is 150MPa, and that internal pressure is the same as Flagstaff Az (75KPa), that means that the workable diameter is 100m.

Let us ignore that for the moment, and assume we want to make a section of cylinder about the same diameter it is long.

For 40 panels, you can make a cylinder 30m in diameter, and 24m long.

The structural loading on this is about 15% of yield, meaning the joining can have quite large stress concentrations without issue.
Ten launches get you the shell of a large toroidal station with the outside diameter 80m, with a cylinder diameter of 30m.
Or a cylinder 200m or so long by 30.

It would need a whole full load of liquid air to fill it to STP, around 150 tons.

You then begin to outfit the inside, once in STP.

(toroidal station of course would not have the segments quite square, and they would be individual.)

Joins would be handled with two million five hundred thousand M8 bolts, all alone in the night.

In-space construction is low TRL.
I'm trying to keep the design to ISS-level module-docking type construction, hence the 8 pie-wedge payloads design.

[MikeAtkinson]

Another way would be to use inflatable structures. Central core + inflatable shell could probably reach 20m diameter, perhaps as much as 30m. About 8m long.

By joining them end-to-end it is possible to create cylinders, rings, etc. Joints could be simple berthing or include a node.

One obvious design would be a ring, with 4 nodes, from those nodes a link is made to a central cylinder made of several more inflatables.

[drzerg]

you can pack any segmented torus inside a cylinder. just make enough small blocks that can be reassembled on orbit

[Lar]

I see the how. I don't see the why.

Large constructions will need to be built in space using asteroidal materials, as to do otherwise implies a LOT of launches. Better to work on developing the mining and refining and transport equipment, IMHO.

But that's not the OP's question. I think the answer is yes, it could be done.

[speedevil]

In-space construction is low TRL.
I'm trying to keep the design to ISS-level module-docking type construction, hence the 8 pie-wedge payloads design.

I do not disagree with this, but to a degree, it's because it's been designed as a complex fragile mass-reduced thing.

Do not make a complex multipurpose robot if a simple one will work.
Yes, the Canadarm is awesome.

However, for the price of the Canadarm, you can buy many, many industrial robots.

The TRL of a general purpose robot able to do up arbitrary fasteners rapidly over a large work envelope is not high.

The TRL of a special purpose  robot able to do up one size of fastener, all in one orientation and bolt pattern, flat against a surface, where it doesn't really matter if all of the fasteners are fixed properly due to massive redundancy is rather higher.
(to join adjacent segments)

The TRL of a simple drawer-slide like mechanism to extend and hold the segment to be bolted on is again quite high, especially if initially you don't care about speed, and are willing to use massive redundancy. (cylindrical is a lot simpler than toroidal, the end-caps would be done after some learning had been done)

Industrial robots are quite capable of most of the tasks - the problems are in the control. This is greatly mitigated if you accept that everything goes very slowly for the initial fixings.
Light modifications to the arms will need to be done for thermal control.

On asteroidal mining - well...
The technical readiness of that is considerably below that of any in-space construction. The bare shell of the ~30m*240m*25mm cylinder above could be purchased for $3m in aluminium cost.
It will cost considerably more to have the segments fabricated into appropriate near-identical shapes, but this is something your average shipyard can do.

I can write a contract, and get people working on it today, for a remarkably small amount of money if I don't tell them it's for space.

Once you have the pressurised volume that the ITS can dock with, start to bring up lightly tested industrial and other equipment and start trying stuff.

[TrevorMonty]

I'm fan Oneil Cylinder as it provides varying levels of gravity, from 0 in middle to max on   shell. Settlements may need 1g but space tourism would prefer low G eg 0.1-0.3. Why go to space to experience 1g may as well stay at home. Gravity needs to be high enough to make daily living enjoyable without hassles associated with 0g. For long term stay being able to sit down for dinner without it floating off plate is must, along with bathroom activities.
With 0g just meters away in centre of cylinder tourists can have best of both worlds.

Depressurization due haul breach by debris  is not as serious as you think, a large cylinder takes 10-100s minutes to depressure with inch hole. Long enough to reach safe areas in haul while plugging  hole. I would hope anything big enough that can make large life threating holes could be avoided. Haul is likely to be layered with all services between floor and outer haul.

BFR definitely bring launch costs down for contruction, $100-200/kg  is realistic when buying dozens of flights all carry low value items like metal sheets. In space assembly where pieces are welded together by robotic arms is biggest challenge but achievable. Ideally construction of pressure haul wouldn't need any humans in space. Fitting out pressurized haul would need humans but they would be working in shortsleeve envirnoment, with gravity.

[rakaydos]

It sounds like the thread consensus is that in-space construction is a significantly better option than ground assembly and docking.

How much progress is being made on assembler robots? I saw the post on the spiderbot, but is it actually funded? What else is out there?

[speedevil]

It sounds like the thread consensus is that in-space construction is a significantly better option than ground assembly and docking.

How much progress is being made on assembler robots? I saw the post on the spiderbot, but is it actually funded? What else is out there?

Nobody has yet tried to do this as a conventional engineering project, which happens to be in space.
You do not try to design space hardware.
You take industrial hardware and make minimal modifications to it, and try it to see if it works.

You do not even go as far as http://mashable.com/2017/11/02/nasa-updates-international-space-station-printer/ - for example, if launch cost drops enough. The ISS needed a new printer, it actually required as far as I can tell one spring and some sponges.
Instead, it got composite parts, completely reengineered cover, lightened, more fastnings, ...

This made 50 $129 printers (as a lifetime buy) cost likely well over $50K each, probably $1M each, counting the fact that maybe four of them are likely to be used.

Compared to, for example, buying 50 different printers, and just trying them on orbit. (in a ventable cupboard, for the first few tests).

With cheap launch, actually trying stuff on orbit may be significantly cheaper than doing a 'proper' design or even test of earth hardware past first blush screening.

[Lar]

With cheap launch, actually trying stuff on orbit may be significantly cheaper than doing a 'proper' design or even test of earth hardware past first blush screening.

I think this is something that people are really struggling to internalise. But it is hugely important.

[rakaydos]

With cheap launch, actually trying stuff on orbit may be significantly cheaper than doing a 'proper' design or even test of earth hardware past first blush screening.

I think this is something that people are really struggling to internalise. But it is hugely important.

"Failure IS an option."

[John Alan]

The first country or entity that builds an honest to goodness space dock in orbit and offers assy services on orbit will make a fortune.
Just saying... a pressurized in orbit workshop that humans and robots can weld and assemble stuff  from brought up piece parts and chunks... is where long term this needs to go... 

[envy887]

Another way would be to use inflatable structures. Central core + inflatable shell could probably reach 20m diameter, perhaps as much as 30m. About 8m long.

By joining them end-to-end it is possible to create cylinders, rings, etc. Joints could be simple berthing or include a node.

One obvious design would be a ring, with 4 nodes, from those nodes a link is made to a central cylinder made of several more inflatables.

Inflatables can be much, much larger than this. If you use 1 mm thick Kevlar sheets you can make a 100 meter diameter sphere, holding 524,000 cubic meters of air at STP (almost 600 times larger than ISS). Double the wall thickness for a safety factor of two and it still only masses ~90 tonnes, and folded down and vacuum packed needs less than 100 cubic meters of launch volume - easily launched on one BFS.

This approach could yield huge pressurized volumes that could be internally fitted out by crew or robots. With such huge volumes depressurization is an unlikely emergency, since even a huge 1 meter hole would take over an hour to depressurize that volume.

IMO, the approach of first inflating a pressure vessel, then constructing usable space inside it, is far easier than making everything modular and docking or doing EVA for construction. Obviously the inflatable needs a bus for a launch mount, visiting vehicle docking, an airlock to the interior, propulsion, comms, some power and thermal control, and to carry the initial inflation charge.

I'm fan Oneil Cylinder as it provides varying levels of gravity, from 0 in middle to max on   shell. Settlements may need 1g but space tourism would prefer low G eg 0.1-0.3. Why go to space to experience 1g may as well stay at home. Gravity needs to be high enough to make daily living enjoyable without hassles associated with 0g. For long term stay being able to sit down for dinner without it floating off plate is must, along with bathroom activities.
With 0g just meters away in centre of cylinder tourists can have best of both worlds.

Depressurization due haul breach by debris  is not as serious as you think, a large cylinder takes 10-100s minutes to depressure with inch hole. Long enough to reach safe areas in haul while plugging  hole. I would hope anything big enough that can make large life threating holes could be avoided. Haul is likely to be layered with all services between floor and outer haul.

BFR definitely bring launch costs down for contruction, $100-200/kg  is realistic when buying dozens of flights all carry low value items like metal sheets. In space assembly where pieces are welded together by robotic arms is biggest challenge but achievable. Ideally construction of pressure haul wouldn't need any humans in space. Fitting out pressurized haul would need humans but they would be working in shortsleeve envirnoment, with gravity.

This, but "construction" of the pressure hull is as simple as unpacking and inflating it. Kevlar has the advantage of being extremely cut and puncture resistant, and if built in 2 to 4 multiple inflatable wall layers (e.g. 1 mm of Kevlar / 10 cm of air / 1 mm Kevlar / repeat) works as a MMOD shield. A thin sheet of Mylar on the exterior could be reflective insulation. Thin-cell solar panels with wiring or flex circuits could be stitched to the outside and folded in with the shell before launch.

Some type of active thermal control like a variable albedo coating or a movable sunshade/reflector would be nice for controlling internal temperatures, although the thermal inertial of 600 tonnes of air helps a lot with smoothing out short term insolation variations.

[rakaydos]

IMO, the approach of first inflating a pressure vessel, then constructing usable space inside it, is far easier than making everything modular and docking or doing EVA for construction. Obviously the inflatable needs a bus for a launch mount, visiting vehicle docking, an airlock to the interior, propulsion, comms, some power and thermal control, and to carry the initial inflation charge.

Also stationkeeping and reactionwheel spindown thrusters. If the raptorRCS from the BFR becomes industry standard, so will the BFS "refueling port"docking standard,, which is another thing to include.

[TrevorMonty]

It sounds like the thread consensus is that in-space construction is a significantly better option than ground assembly and docking.

How much progress is being made on assembler robots? I saw the post on the spiderbot, but is it actually funded? What else is out there?

The current develops for in space construction are printing/assemble large light weight trusses or boom structures. This are for antennas, sunshades and solar arrays, persistent platforms. Besides building structure the robots will fit them out with cloth, solar panels, wiring etc.

I haven't heard about anybody working on systems to assemble large pressure vessels in space. This in way maybe easier as it would use precut metal parts that just need welding or fastening together. At present there is no market for it except on ground. As it happen the companies with lot of experience in this type of construction is LV manufacturers, look at how they stir weld SLS tanks together.

[Coastal Ron]

Some assembly required.

Unless you design your rotating station to be built using modules that are sized the transportation being used, some assembly (or even a lot) will be required in the hard vacuum of space. Not that it's bad to do that, just to point that out so that no one is scared about it - we did assembly on the ISS and it worked out OK.

Just as a suggestion, while the BFR/ITS is supposed to be 9m in diameter, and there is a cargo version that Musk showed off, designing your station to exclusively use the BFR/ITS may not be the best strategy. Could you build what you want if you designed everything to fit in the narrower Blue Origin New Glenn? It offers something close to 7m in diameter, and possibly much longer sections.

That way you would have two transportation systems that you could use, and that would ensure that you don't stop work if one of them shuts down their service for some reason (or doesn't want your business).

[TrevorMonty]

Thought bit more on how this construction could be done. Have curve platform  with robot arm with stir welder, bolt or clamp two section to platform then weld them together repeat till have complete cylinder section/ring. Now bolt or clamp two cylinder sections/rings together and stir weld the join. The one small welding platform which only needs to be big enough to hold two sections together, could assemble massive cylinders.

The sections used to create cylinder would all be identical allowing for mass production on ground or in distant future moon and asteriod mines.

Edited after reading Costal comments. Section sizes are designed to maximize LV lift capabilies. Mixing sizes is not big issue, one ring can be made from 20 sections while another could be 30 smaller sections, as long as same diameter doesn't matter. Ring width can be different, eg use 3 x  5m x30 section rings or 2 x 7.5m x 20 sections to create same 15m long cylinder.

[watermod]

I recall some long ago "research".   Maybe SBIR/STTR or the like that explored a steered and propelled device that extruded molten or glassy or plastic material on a structural flight path to provide girders or surfaces for a space structure.   Always liked the concept.

First example had some material like thermite melting aluminum or iron and extruding it in a stream while steering the extruding device in space. 

[Nomadd]

 BA-2100 concept. Each module about 8m stowed and around 100 tons if fully equipped, 70 tons dry weight.

[punder]

How do you dock with it?

Two hab modules, a docking hub, and a truss would make a "minimal" rotating facility with two levels of g, given an offset center of gravity. In my pic, the blue hab is for Lunar G, and the red hab is Mars g. (No, the distances aren't accurate, and I didn't include solar arrays, radiators and whatnot, but that's what you get for five minutes of work.)

Here is an interesting link: http://www.quantumworks.com/jbis_article.htm

[lamontagne]

A complex space build station, and a simpler assembled station, made from expandable units and extendable panels

[Lar]

I like the skylon in the second render...

[punder]

A complex space build station, and a simpler assembled station, made from expandable units and extendable panels

I bow down before you, sir.   

But my interest is in rotating research stations that can be built near-term. I just can't believe it's that difficult. And we only have two data points for the biological effects of gravity equal to or less than 1g: 1g and 0g. A simple station as I described would pretty much, uh, revolutionize our knowledge on this topic, which is absolutely critical for human expansion into the Solar System.

Note that many values of g are possible with such a station because 1) each hab has multiple levels, and 2) angular velocity can be changed at will using thrusters.

[Oersted]

How do you dock with it?

Two hab modules, a docking hub, and a truss would make a "minimal" rotating facility with two levels of g, given an offset center of gravity. In my pic, the blue hab is for Lunar G, and the red hab is Mars g. (No, the distances aren't accurate, and I didn't include solar arrays, radiators and whatnot, but that's what you get for five minutes of work.)

Here is an interesting link: http://www.quantumworks.com/jbis_article.htm

Punder, that graphic in your posting also suggests a new kind of docking, the "screw it in" method!  :-)

[punder]

How do you dock with it?

Two hab modules, a docking hub, and a truss would make a "minimal" rotating facility with two levels of g, given an offset center of gravity. In my pic, the blue hab is for Lunar G, and the red hab is Mars g. (No, the distances aren't accurate, and I didn't include solar arrays, radiators and whatnot, but that's what you get for five minutes of work.)

Here is an interesting link: http://www.quantumworks.com/jbis_article.htm

Punder, that graphic in your posting also suggests a new kind of docking, the "screw it in" method!  :-)

Ha, hey I'm just channeling Stanley Kubrick.      Just match the roll rate, no problemo.

[Oli]

The hab in the Martian looked pretty good to me. Low g and large windows for a great view. I guess the interior could be a bit more comfy.

[Steve D]

We are talking a lot about what will fit in the BFR. What about how we will be getting these things out the door? We have a 9 meter diameter space but will the door allow a 9 meter object to pass through it?

[Barrie]

I recall some long ago "research".   Maybe SBIR/STTR or the like that explored a steered and propelled device that extruded molten or glassy or plastic material on a structural flight path to provide girders or surfaces for a space structure.   Always liked the concept.

First example had some material like thermite melting aluminum or iron and extruding it in a stream while steering the extruding device in space.

Like large-scale 3D printing using a free-flying print-head?

[speedevil]

We are talking a lot about what will fit in the BFR. What about how we will be getting these things out the door? We have a 9 meter diameter space but will the door allow a 9 meter object to pass through it?

The BFS has a cargo hatch that appears to be capable of passing a ~9m*3.6*3.6m object (or a hair more in width and height), and you can in principle stack several such items on a rail system for deployment out of the hatch. (if you remove the cabins, or as I have hypothesised, they're on rails for easy removal)

The BFC is less well defined, and has never had what seems like more than a notional sketch.
Similarly, the final version of the tanker, other than it will look 'silly'.

As one point, if the cabins of BFS are in fact demountable cargo containers, they fit easily through the hatch, and can have mating connectors on the faces, the problem almost resolves to one of Lego.
They are perhaps smaller than one would like, but ready modularity may trump that, especially as they would in principle be easily manufacturable by any manufacturer globally, and shipped readily in standard transport.

[watermod]

I recall some long ago "research".   Maybe SBIR/STTR or the like that explored a steered and propelled device that extruded molten or glassy or plastic material on a structural flight path to provide girders or surfaces for a space structure.   Always liked the concept.

First example had some material like thermite melting aluminum or iron and extruding it in a stream while steering the extruding device in space.

Like large-scale 3D printing using a free-flying print-head?

Yes!

[oldAtlas_Eguy]

I recall some long ago "research".   Maybe SBIR/STTR or the like that explored a steered and propelled device that extruded molten or glassy or plastic material on a structural flight path to provide girders or surfaces for a space structure.   Always liked the concept.

First example had some material like thermite melting aluminum or iron and extruding it in a stream while steering the extruding device in space.

Like large-scale 3D printing using a free-flying print-head?

Yes!

A device that can print aluminum can print a seamless torus with any openings with flanges and mounting point all integrated to the primary torus structure. Only needs the blanks/ingots to feed the 3D printer. The printer has its own power solar array as part of the device plus arms that can stabilize the device while printing to the item that is being printed. Grab points are printed as the printer prints the torus. No limitations on size. Just takes longer a linear calculation based on thickness and area of wall.

This is the end direction current research is going. All other in space assembly is at best a temporary measure or a supporting capability to have multiple 3D printers printing many parts simultaneously with the assembly of the parts printed. Increases the construction completion time from start to finish. A BTW decking is printed as well so no need to install any structural elements inside the torus. Floors, walls, stairs, mounts, cabinets, etc are all printed simultaneous to the printing of the shell.

Assumption of a project:
1 - First launch is the 3D printer. This device is not a mass constraint but a volume constraint due to its solar arrays, thermal radiators, arms, on-board spare parts (such as print heads and even arms), propulsion system, and the 3D printer.

2 - Next are the launches of the aluminum blanks. These launches occur as often as the 3D printer can use up the blanks.

3 - Next is the launch of the internal and external parts/equipment installation robots. External assembly robots have solar arrays and propulsion.

4 - Next is the non-printable internal and external parts. Priority is to power generation, thermal control systems, and the distribution of power, cooling. Next is instruments and computer/controls and comm distributions (Ethernet or Ethernet like systems). Next is other habitation items. Water, farming, ECLSS, furniture, displays(Monitors) and human interfaces.

5 - Human habitation supplies.  After sufficient pressurants the the torus is pressurized then the torus systems is fueled for initial spin-up. Torus is now available for habitation with the arrival of pressurized supplies.


Costs are somewhat linear but actually get a little cheaper per unit volume as total volume increases.

[Barrie]

Any sense in having a coarse, high deposition rate print-head for structure, and smaller high-res print-heads that add precision details such as mounting lugs for attached equipment?

[oldAtlas_Eguy]

Any sense in having a coarse, high deposition rate print-head for structure, and smaller high-res print-heads that add precision details such as mounting lugs for attached equipment?

If it is the exact same materiel then there is no reason not to. The course system deposits most of the materiel while the fine system adds smooth surfaces and other fine feature items.

These systems would be operating 24/7 continuously as long as there is a supply of blanks. A robot that transports stores blanks and can shift them over to a printer may be another robot needed to speed up construction. It would accept a specialized container of the blanks such that it would hold 2 of them so that when a BFR delivers a new container of blanks the empty one is offloaded to the BFR for return to Earth.

Eventually the Blanks will stop coming from Earth but arrive from a IRU plant receiving raw material from either the Moon or an asteroid. at which point the BFR does not slow down in flights but switches to the deliver of the everything else: robots, accouterments, supplies, and humans.

[KelvinZero]

Shorter term, for minimum assembly:

What about a telescoping nested set of cylinders? The thing expands to a long truncated cone instead of a long cylinder, but so what? This would fit arbitrarily well inside the Cargo BFS bay, because the inner cylinders could be longer than the outer ones.

Each truncated cone could be engineered with a slight curve so that a handful of them could fit together to form a torus. The tube radius would have bulges approaching 9m but go down to say 3m. Straight versions of the truncated cones could be used to reach the hub where docking takes place.

[MikeAtkinson]

Building the structure, whether by 3D printing, bolting together plates, or other means, is only solving the easiest problem.

After the structure is completed then it needs to be kitted out with mechanics, data and power cabling,plumbing, life support and many other systems. Then there is the QA, testing, faultfinding and repair. All much more expensive and time consuming.

[TrevorMonty]

Shorter term, for minimum assembly:

What about a telescoping nested set of cylinders? The thing expands to a long truncated cone instead of a long cylinder, but so what? This would fit arbitrarily well inside the Cargo BFS bay, because the inner cylinders could be longer than the outer ones.

Each truncated cone could be engineered with a slight curve so that a handful of them could fit together to form a torus. The tube radius would have bulges approaching 9m but go down to say 3m. Straight versions of the truncated cones could be used to reach the hub where docking takes place.

The whole idea of Oneil Cylinders is to have large enough diameter to create artifical gravity when rotating around 2-4rpm. To get a useful gravity even 0.1g the diameter needs to be 10s of meters.

[TrevorMonty]

3D printing has its place but printing 100m3 of alloy plate strong enough to hold breathable atomsphere is not realistic. I not sure 3d print can match high tensile plate manufactured by tradition methods, if not then printed structure will be heavier.

[TrevorMonty]

Building the structure, whether by 3D printing, bolting together plates, or other means, is only solving the easiest problem.

After the structure is completed then it needs to be kitted out with mechanics, data and power cabling,plumbing, life support and many other systems. Then there is the QA, testing, faultfinding and repair. All much more expensive and time consuming.

Nanoracks and their partners are trying to do just this with Centuar US, they think it can be done robotically.

If it works out then maybe Oneil Cylinders 10s meters in diameter are possible and affordable in near future using earth launched materials.

[KelvinZero]

The whole idea of Oneil Cylinders is to have large enough diameter to create artifical gravity when rotating around 2-4rpm. To get a useful gravity even 0.1g the diameter needs to be 10s of meters.

To clarify, I was discussing a torus with 9m only referring to the tube radius.

[Nomadd]

To clarify, I was discussing a torus with 9m only referring to the tube radius.

I've always thought of that with smaller, concentric rings inside, or large enough spokes to have living and work areas at varying gravities for rehab or research.

[lamontagne]

Here is a station design more appropriate for this thread.

The modules are 8m in diameter and all brought up by BFS.  Is is an assembled station, I have doubts 3D printing will be fast enough at first.  Probably eventually though, then larger stations will be possible.

The two docking ports can be brought down to 0 RPM for docking, but spend most of their time turning with the station.  This reduces the wear and tear on the rotating joint to a minimum.
This version is 60 m in radius, so 0,5g at 3 rpm. 0r 1/4g at 2 rpm.
At this stage there are only 4 modules, but eventually the ring could be completed.  Each module is separated from its neighbor by an airlock.  There are only 2 tubes to the center  this early in construction. As the station is built 2 more tubes could be added.
The structure is mainly thin cables that work in tension, attached to central rings, very close to a bicycle structure.  The 4 large spokes are structural and work in compression with added wires, not shown, and serve to dissipate vibration energy.  Lots of wires.
I must admit I worry about collision for this type of design, that is why I extended the ports far from the rotating hub.
Could modules be added while the station is in rotation?  Seems unsafe, so I expect they would be added by batches, after a few were accumulated in orbit nearby.

[blasphemer]

Bigelow BA 2100 is around 100 tons (some sources say 65 tons), reusable BFR can lift 150 tons. Maybe a Bigelow module built specifically for BFR will make sense.

BA 5000? 

[lamontagne]

Bigelow BA 2100 is around 100 tons (some sources say 65 tons), reusable BFR can lift 150 tons. Maybe a Bigelow module built specifically for BFR will make sense.

BA 5000? 

More likely Bigelowe 1200.  They are rated by internal volume and the expansion factor is about 1.5.  Musk stated about 850m3 internal volume for BFS.

The Bigelowe module is much heavier than 16 tonnes because is has much thicker walls and includes all the fittings.  My mass is for a simple carbon fiber shell, with little else.  Outfitted, it would be much heavier.

[lamontagne]

Complete station.  48 modules.  About 36 000 m3 of space. 

[speedevil]

Building the structure, whether by 3D printing, bolting together plates, or other means, is only solving the easiest problem.

After the structure is completed then it needs to be kitted out with mechanics, data and power cabling,plumbing, life support and many other systems. Then there is the QA, testing, faultfinding and repair. All much more expensive and time consuming.

'All much more expensive'.

Hmm.
Taking the somewhat extreme case that we have a dark pressure vessel with O2/N2 in it, indifferent thermal control, and that we are not going to use it other than as an emergency backup, what are the costs of outfitting?

You need a rebreather, (1) and a gas-tight suit (2). Neither is life-critical, as it only needs to work 99% or so, in order to reduce the offgassing from the humans to easily managable levels. Add dessicant to take 1kg of water. (

You need a thermal managment package (3) in case you get a bit warm, and you need a head-torch and portable lights.
Supplies for the above of course, and you have a functioning human that may grumble a bit about its suit, but is quite able to work for 8h/day.

Cost $10K or so for initial outfit, and $2K for launch.
$750 a day for supplies, $900 a day with one-use rechargeable batteries, with some left over for tools and stuff.

Assume a three week life for all of the outfit, and that's $1500 approximately per day, or $200/hr.

$200/hr is the initial price before we have any lights, power, and when we are actually throwing overboard (or deorbiting) ice after unfreezing it for thermal control, and rechargeable batteries after using them once.

A simple spinning drum, with a hose to vacuum which you put the water in with ice-cube trays around the outside will drastically reduce ice shipments, dropping it by $300/day to $1200/day.

Similary, a cupboard with a simple water-heat-exchange loop going from the inside to outside and many closely fitting shelves to place the bags of dessicant in, with a hose to vacuum drops that from $50/day to $0 or so. Minor repairs on the suit with duct tape, another $75/day.
$1075/day.

Notice that the rebreather is quite expensive, and get it serviced (on earth) and shipped back for $1000/21 days.
$750/day.

This is beyond the point that you can very easily start trading labour for complex machinery.
It becomes literally possible to employ someone to go round to everyone and swap out their cannisters, and duct-tape up any small leaks in their suits.

At this point, you actually have to start counting the pay of the astronaut, as it gets significant. ($200-300/day or so).

If you have rebreathers, and you can use them in this manner worst case, life support development becomes lots less risky.

Llife support development is made much easier by the fact that - even absent the possibility of multiple crew able to work on it and fix it, and the fact it's mass unconstrained, you can ship it back to earth almost free.
 
I am not quite saying that life-support should be made from stuff from home depot.
But you actually pretty much could.

1) https://www.oceanenterprises.com/scuba-gear/poseidon-mkvi-discovery-rebreather-system-en.html $7K
2) https://www.allpipe.co.uk/products/detail/heavy-duty-chemical-suits/trellchem-freeflow-chemical-suit ($1500)
3) http://www.veskimo.com/why-best-cooling-vest.php ($1000)
4) wherever, $100
5) https://www.shootingandscuba.co.uk/store/8-pack-co2-absorber-cartridge-for-poseidon-mkvi-from-tal-shootingandscuba $200/24h.
6) Ice, 50kg a day. ($500, assuming it is all disposed of after one use)
7) Oxygen 1kg a day, $10.
Silica gel dessicant, 3kg to 30% weight in water. $50/day. https://www.ebay.co.uk/itm/3-8L-Kitty-Sand-Silica-Gel-Cat-Litter-Buy-1-3-or-8-Bags-Value-Bulk-Deal-Natural/231510190134

[mikelepage]

So the BFR will offer a whole new level of payload to orbit. Is it enough to build a pre-assembled Oneill Colony? Spin gravity and all?

I hate to be a pedant, but an "O'neill cylinder" very specifically refers to a cylinder of 8km diameter and 32km length: So the answer is a pretty simple no.
https://en.wikipedia.org/wiki/O%27Neill_cylinder

If you're talking about using BFR to deliver smaller scale spin-gravity habitats to orbit, I whole-heartedly agree, but MODS can we please rename the thread?     Loose use of terminology does no one any favours.

[QuantumG]

If you're talking about using BFR to deliver smaller scale spin-gravity habitats to orbit, I whole-heartedly agree, but MODS can we please rename the thread?     Loose use of terminology does no one any favours.

Changing the thread title on the first post is all you need to do to rename the thread.

[mikelepage]

Complete station.  48 modules.  About 36 000 m3 of space.

Nice work.  Hopefully you'll put it in a higher orbit than that or else you'll lose it to atmospheric drag before you finish building it (just being cheeky - I know it's a concept ) For something that big you'd want to be up around the 800km altitude range, or else you'd spend a ridiculous amount of fuel reboosting it.

This is a graph of the lifetime - until drag brings it down into the atmosphere - versus altitude.  (Van Allen Belts start around 1000km)

[KelvinZero]

Complete station.  48 modules.  About 36 000 m3 of space.

Hi is that based on https://en.wikipedia.org/wiki/BA_2100 or similar?

It looks like you are putting new doors through the inflatable section and including a sizeable new cylindrical join to that section.

IMO you would use the two doors that come at each end of the BA2100, connected to the central rigid part. You would only need a small adaption to one end so that when they joined end to end they formed a curve. Cables to the center could connect to this adaption also.

[rakaydos]

So the BFR will offer a whole new level of payload to orbit. Is it enough to build a pre-assembled Oneill Colony? Spin gravity and all?

I hate to be a pedant, but an "O'neill cylinder" very specifically refers to a cylinder of 8km diameter and 32km length: So the answer is a pretty simple no.
https://en.wikipedia.org/wiki/O%27Neill_cylinder

If you're talking about using BFR to deliver smaller scale spin-gravity habitats to orbit, I whole-heartedly agree, but MODS can we please rename the thread?     Loose use of terminology does no one any favours.

Pedantry acknowledged. :p

[rakaydos]

Complete station.  48 modules.  About 36 000 m3 of space.

Hi is that based on https://en.wikipedia.org/wiki/BA_2100 or similar?

It looks like you are putting new doors through the inflatable section and including a sizeable new cylindrical join to that section.

IMO you would use the two doors that come at each end of the BA2100, connected to the central rigid part. You would only need a small adaption to one end so that when they joined end to end they formed a curve. Cables to the center could connect to this adaption also.

I doubt BE inflatables are stressed to handle 1g (or even 1/3g) while inflated, certiantly not on a permanant basis with people moving around in them.
I would expect it would require a new design.

[DreamyPickle]

There is an older thread for a "realistic near-term rotating space station".

One proposal to simplify docking was to have just two ports in the center and put vehicles in a constant roll as they dock. This requires the center of mass of the visiting spacecraft to be aligned with the docking port and will very likely be false for the BFS. Most other craft dock with their nose but for BFS this would require a flexible cover in the most sensitive area of the heat shield.

In order for the BFS to dock the station would need a constantly rotating joint around a central non-rotating docking spike. Smaller vehicles could dock on the sides of the spike but the BFS would either have to dock in the plane of rotation (2 ports only) or otherwise the spike would have to be extend at least ~10-15 meters away from any rotating spokes.

[rakaydos]

In order for the BFS to dock the station would need a constantly rotating joint around a central non-rotating docking spike. Smaller vehicles could dock on the sides of the spike but the BFS would either have to dock in the plane of rotation (2 ports only) or otherwise the spike would have to be extend at least ~10-15 meters away from any rotating spokes.

What about an airlock/"spinlock" combo?

That is, a module that can alternate between loosely bearing'd to the station, non spinning and attached to a non-spinning spacecraft with an airtight seal... to only loosely bound to the non-spinning spacecraft, spinning with the space station, with an airtight seal with the station?

It would  bypass the usual problems of making an airtight bearing with minimal friction, because it is spun up and down each time it's used, and only sealed with the side it matches rotation with.

[KelvinZero]

I doubt BE inflatables are stressed to handle 1g (or even 1/3g) while inflated, certiantly not on a permanant basis with people moving around in them.
I would expect it would require a new design.

Nah, just google a few pictures of it being used, on the moon and so on. This is exactly how it is meant to be used. This orientation is fine. All forces handled through the two rigid ends is exactly what they are designed for.

When it is sitting on a test stand on earth, under one gravity, that is exactly the final stresses it is intended for. No hair raising difficult to test optimisation of a few pounds. A hugely more convenient scenario.

[rakaydos]

There is an older thread for a "realistic near-term rotating space station".

blerg... I got halfway through the thread before I got fed up with the arguments about NASA, artificial gravity research, and mars programs.

The earlier argument about needing a rigid (compressed) component to dampen oscillation in a tensioned tether was at least useful.

[mikelepage]

There is an older thread for a "realistic near-term rotating space station".

blerg... I got halfway through the thread before I got fed up with the arguments about NASA, artificial gravity research, and mars programs.

The earlier argument about needing a rigid (compressed) component to dampen oscillation in a tensioned tether was at least useful.

I'm happy to let the older threads die, since BFR seems like a much more likely vehicle (and SpaceX a much more likely partner company) to facilitate spin-gravity habitat research.  I've been somewhat quieter of late because I've been working on Exodus Space Systems, the Australian space startup company I co-founded to deal with exactly these problems.

I'm attaching here an excerpt from our vision/strategy document, which I think provides a good summary of the various considerations that spin gravity proposals should take into account.  Hopefully it should be useful to some of the newer participants, but I'd also appreciate feedback - particularly if you think I've missed anything.

[speedevil]

BFR cost to orbit - about $10000/ton. (to make the 'economy ticket' comment for passenger transport work).

SLS program cost - ~$22B out to 2024.

Two million five hundred thousand tons of spinning metal.
There is only one possible station design.

[livingjw]

There is an older thread for a "realistic near-term rotating space station".

blerg... I got halfway through the thread before I got fed up with the arguments about NASA, artificial gravity research, and mars programs.

The earlier argument about needing a rigid (compressed) component to dampen oscillation in a tensioned tether was at least useful.

I'm happy to let the older threads die, since BFR seems like a much more likely vehicle (and SpaceX a much more likely partner company) to facilitate spin-gravity habitat research.  I've been somewhat quieter of late because I've been working on Exodus Space Systems, the Australian space startup company I co-founded to deal with exactly these problems.

I'm attaching here an excerpt from our vision/strategy document, which I think provides a good summary of the various considerations that spin gravity proposals should take into account.  Hopefully it should be useful to some of the newer participants, but I'd also appreciate feedback - particularly if you think I've missed anything.

One order of magnitude reduction in cost can be expected in the next 5-10 years. Not multiple orders of magnitude. 2 at most a long time from now.

John

[rakaydos]

There is an older thread for a "realistic near-term rotating space station".

blerg... I got halfway through the thread before I got fed up with the arguments about NASA, artificial gravity research, and mars programs.

The earlier argument about needing a rigid (compressed) component to dampen oscillation in a tensioned tether was at least useful.

I'm happy to let the older threads die, since BFR seems like a much more likely vehicle (and SpaceX a much more likely partner company) to facilitate spin-gravity habitat research.  I've been somewhat quieter of late because I've been working on Exodus Space Systems, the Australian space startup company I co-founded to deal with exactly these problems.

I'm attaching here an excerpt from our vision/strategy document, which I think provides a good summary of the various considerations that spin gravity proposals should take into account.  Hopefully it should be useful to some of the newer participants, but I'd also appreciate feedback - particularly if you think I've missed anything.

One order of magnitude reduction in cost can be expected in the next 5-10 years. Not multiple orders of magnitude. 2 at most a long time from now.

John

Isnt Facon 9 reuse a reduction of costs (not nessesarally prices) of an order of magnitude? And BFR supposed to be another order of magnitude?

[speedevil]

One order of magnitude reduction in cost can be expected in the next 5-10 years. Not multiple orders of magnitude. 2 at most a long time from now.

I assume this means you don't believe that passenger transport at close to the stated goals will come about ever?

[livingjw]

One order of magnitude reduction in cost can be expected in the next 5-10 years. Not multiple orders of magnitude. 2 at most a long time from now.

I assume this means you don't believe that passenger transport at close to the stated goals will come about ever?

Maybe, a long time from now.  Going from $60,000,000 to $6,000,000 to $600,000 will take time.  That's all I'm saying.

John

[lamontagne]

Complete station.  48 modules.  About 36 000 m3 of space.

Hi is that based on https://en.wikipedia.org/wiki/BA_2100 or similar?

It looks like you are putting new doors through the inflatable section and including a sizeable new cylindrical join to that section.

IMO you would use the two doors that come at each end of the BA2100, connected to the central rigid part. You would only need a small adaption to one end so that when they joined end to end they formed a curve. Cables to the center could connect to this adaption also.

No its based on a rigid module optimized to fit inside the BFS fairing.

[lamontagne]

Complete station.  48 modules.  About 36 000 m3 of space.

Hi is that based on https://en.wikipedia.org/wiki/BA_2100 or similar?

It looks like you are putting new doors through the inflatable section and including a sizeable new cylindrical join to that section.

IMO you would use the two doors that come at each end of the BA2100, connected to the central rigid part. You would only need a small adaption to one end so that when they joined end to end they formed a curve. Cables to the center could connect to this adaption also.

This was my inflatable solution  :-)  start with 3 modules, then add and add and add.....

[mikelepage]

Complete station.  48 modules.  About 36 000 m3 of space.

Hi is that based on https://en.wikipedia.org/wiki/BA_2100 or similar?

It looks like you are putting new doors through the inflatable section and including a sizeable new cylindrical join to that section.

IMO you would use the two doors that come at each end of the BA2100, connected to the central rigid part. You would only need a small adaption to one end so that when they joined end to end they formed a curve. Cables to the center could connect to this adaption also.

No its based on a rigid module optimized to fit inside the BFS fairing.

Hi Lamontagne, great work, but I'm curious how you figured out how big the BFS payload bay was?

By eyeball it all looks right to me, but Elon did say at 18 minutes into the talk that "that payload bay is eight stories tall, and you could fit a whole stack of Falcon 1s in there".  Looking it up, the Falcon 1 is listed as 21.3m tall. 

In any case, I've been thinking the payload bay must 22-24m long/tall, or just under half the length of the vehicle (48m).  Just wondering how you square that?

EDIT: just looking at your diagrams a little closer and noticed the smaller nose cone section with the little person in there. I suppose you would be able to squeeze in a falcon 1 if you took it right to the tip of the nose.  Makes me wonder if there will ever be a small pressurised section on "cargo" BFSs.

[KelvinZero]

I was just thinking about estimating some values for my telescoping idea.

I think I can work out the basic mass per volume from something like this
https://en.wikipedia.org/wiki/Pressure_vessel#Cylindrical_vessel_with_hemispherical_ends

What is a sensible, non Sci-Fi material choice? What should I use for density and maximum working stress? What should I use for pressure?

Im not asking just for numbers from a table, but also for example what safety margin is it typical to engineer for. Im guessing you don't just engineer for exactly one atmosphere for example. Im guessing those formulae are the theoretical point at which the vessel bursts, ie do not include predefined safety margins. I was thinking there might be some specific guidelines for space, eg engineer for 3x the pressure or such.

[speedevil]

I was just thinking about estimating some values for my telescoping idea.

I think I can work out the basic mass per volume from something like this
https://en.wikipedia.org/wiki/Pressure_vessel#Cylindrical_vessel_with_hemispherical_ends

What is a sensible, non Sci-Fi material choice? What should I use for density and maximum working stress? What should I use for pressure?

Even using quite modest materials allows really ridiculous volumes.
For example, aluminum as used in scaffolding poles can support ~290MPa of stress.
A 1mm skin can in theory support around 290MPa/75kPa = 3900 times its area = 3.9m. (75kPa chosen as it is the pressure at the first highest city in the USA I'd heard of, Flagstaff).
So, a 1mm skin can (just) support an 8m diameter cylindrical atmospheric load, only considering one axis.

You probably don't in fact want to work close to the limit, and 75% is a sane maximum ever load. Even really, really large structural margins lead to quite large structures. Consider a nesting series of cylinders 8.5m or so in diameter and 8m or so long, 1" thick, for example.
With 150 tons of launch, you get 60m^3 of aluminium, or enough to make a one inch approximately cylindrical structure 8m diameter and 100m long.  (three times the B2100), for something you can order inexpensively constructed tomorrow, with rapid delivery.

This is with a naive safety factor of 25 or so.
Bolting this together will get you at worst 50% safe working cyclic load, with a factor of 12 safety.

The proper design needs to consider launch costs, and if you're at more than double launch costs for your entire construction, you may not be doing things cheaply enough.

It is reasonably arguable that you should for a system like BFR, even for initial launches, you should be aiming to design not for todays launch cost, but tomorrows.

The volume and mass potentially available on orbit means that multiple approaches can be tried on orbit at once, inexpensively.
You don't spend seven years and five billion dollars developing robot arms, you go to the market, and buy the twenty most popular robotic arms for a million or so, lightly modify them so they can probably cope with a vacuum, and try them in orbit.

I am not saying that complex payloads are bad, but that we need to look carefully at the environment that shaped the payloads, and consider carefully if that's what we want to do.

The B2100 is a great example of a payload designed for todays launchers.
It is enormously highly engineered and optimised.
But how much would it cost to buy 50 of them.
It's not optimised for a launch cost of $100/kg, never mind $30 or $10.

[tdperk]

One order of magnitude reduction in cost can be expected in the next 5-10 years. Not multiple orders of magnitude. 2 at most a long time from now.

I assume this means you don't believe that passenger transport at close to the stated goals will come about ever?

Maybe, a long time from now.  Going from $60,000,000 to $6,000,000 to $600,000 will take time.  That's all I'm saying.

John

It will take 5 (+5/-0 years tolerance) for costs to enter the <$100/lb range, R&D amort. neglected.  When prices approach that is a different question.

[tdperk]

I was just thinking about estimating some values for my telescoping idea.

I think I can work out the basic mass per volume from something like this
https://en.wikipedia.org/wiki/Pressure_vessel#Cylindrical_vessel_with_hemispherical_ends

What is a sensible, non Sci-Fi material choice? What should I use for density and maximum working stress? What should I use for pressure?

Even using quite modest materials allows really ridiculous volumes.
For example, aluminum as used in scaffolding poles can support ~290MPa of stress.
A 1mm skin can in theory support around 290MPa/75kPa = 3900 times its area = 3.9m. (75kPa chosen as it is the pressure at the first highest city in the USA I'd heard of, Flagstaff).
So, a 1mm skin can (just) support an 8m diameter cylindrical atmospheric load, only considering one axis.

You probably don't in fact want to work close to the limit, and 75% is a sane maximum ever load. Even really, really large structural margins lead to quite large structures. Consider a nesting series of cylinders 8.5m or so in diameter and 8m or so long, 1" thick, for example.
With 150 tons of launch, you get 60m^3 of aluminium, or enough to make a one inch approximately cylindrical structure 8m diameter and 100m long.  (three times the B2100), for something you can order inexpensively constructed tomorrow, with rapid delivery.

This is with a naive safety factor of 25 or so.
Bolting this together will get you at worst 50% safe working cyclic load, with a factor of 12 safety.

The proper design needs to consider launch costs, and if you're at more than double launch costs for your entire construction, you may not be doing things cheaply enough.

It is reasonably arguable that you should for a system like BFR, even for initial launches, you should be aiming to design not for todays launch cost, but tomorrows.

The volume and mass potentially available on orbit means that multiple approaches can be tried on orbit at once, inexpensively.
You don't spend seven years and five billion dollars developing robot arms, you go to the market, and buy the twenty most popular robotic arms for a million or so, lightly modify them so they can probably cope with a vacuum, and try them in orbit.

I am not saying that complex payloads are bad, but that we need to look carefully at the environment that shaped the payloads, and consider carefully if that's what we want to do.

The B2100 is a great example of a payload designed for todays launchers.
It is enormously highly engineered and optimised.
But how much would it cost to buy 50 of them.
It's not optimised for a launch cost of $100/kg, never mind $30 or $10.

Aluminum should be rejected on account of spallation radiation.  Carbon Fiber is a better match  I think.

The 100$ full density a BFR would carry is what, 11lbs/ft^3 ?  Al is 168.5 lbs /ft^3., CF is half that.

Considering the advantage of using standardized size modules (as opposed to telescoping assemblies which I take to be like matryoshka dolls), I think nesting subassemblies which become modules after assembly are the way to go and have a slight edge over BA style expanding modules -- when assembly on orbit becomes assumed.

Until then BA expanders have an advantage of being first day useful.  I believe they will be the places the people maintaining and operating the assemblers live.

[KelvinZero]

Hi, ok I will probably go carbon fibre as a good but non-scifi material that happens to exist in this table:

https://en.wikipedia.org/wiki/Ultimate_tensile_strength#Typical_tensile_strengths
This gives
tensile strength = 1600 for laminates ( I assume the 4137 for fibres alone is just in one direction? )
density = 1.75 g/cm³

I found this inside the link above: https://en.wikipedia.org/wiki/Factor_of_safety

Buildings commonly use a factor of safety of 2.0 for each structural member. The value for buildings is relatively low because the loads are well understood and most structures are redundant. Pressure vessels use 3.5 to 4.0, automobiles use 3.0, and aircraft and spacecraft use 1.2 to 3.0 depending on the application and materials. Ductile, metallic materials tend to use the lower value while brittle materials use the higher values.


( btw, https://en.wikipedia.org/wiki/Basalt_fiber looks really attractive for lunar applications. Strong, fireproof, good elastic strength. No other materials added .. if you are going to do one construction material ISRU on the moon, this sounds great.)

=====

For a carbon fiber cylinder of 8.5m radius and 1atm pressure, I work it out as 50kg/m 12.5kg/m. lets triple it for a lot of safety to 150kg/m 37.5kg/m. A 150 ton BFS cargo could lift 1km ~4km of this cylinder by mass.. which probably means my telescoping idea would have to deal with significant narrowing. I will have to work out the wall thickness and guess the delta width for each segment. It might narrow to an unusable width before it can exploit the BFS launch capability.

=====
edit: appended new working 15/Nov/17

I think this works out to about 6 m³ per kg.

That is a very useful figure because you would get the same result regardless of cylinder radius so it can be applied to all the cylinder sections.

That is about 0.7kg/m² when 8.5m radius. Im getting really thin depth values, under half a millimeter.

That is much thinner than I thought. Does that sound reasonable or have I dropped some decimal places somewhere?

That thin, perhaps you could forget telescoping and send the whole tube rolled up like one of those party whistles.

[mikelepage]

What's the thinking on maintaining pressure in a telescoping tube?

I'm picturing that each end of each tube segment has some kind of inflatable rubber gasket that abuts the next/previous tube (similar to those on the hatch covers in the ISS cupola), but that kind of arrangement wouldn't be able to hold pressure whilst tubes are sliding against each other. 

I guess it's a related problem to that of trying to hold pressure against a rotating bearing (as in a habitat with rotating and stationary parts).  It's easier if you don't try to hold an atmosphere against moving parts, but parts that move can have several set positions which can hold pressure.  Telescoping tubes will only be used in either contracted or extended forms - i.e. you extend it out first, then you bring it up to pressure.

[QuantumG]

Slide rings.

[aero]

They manage to hold pressure in hydraulic rams so I imagine it could be done on a tube.

[KelvinZero]

What's the thinking on maintaining pressure in a telescoping tube?

I'm picturing that each end of each tube segment has some kind of inflatable rubber gasket that abuts the next/previous tube (similar to those on the hatch covers in the ISS cupola), but that kind of arrangement wouldn't be able to hold pressure whilst tubes are sliding against each other. 

I guess it's a related problem to that of trying to hold pressure against a rotating bearing (as in a habitat with rotating and stationary parts).  It's easier if you don't try to hold an atmosphere against moving parts, but parts that move can have several set positions which can hold pressure.  Telescoping tubes will only be used in either contracted or extended forms - i.e. you extend it out first, then you bring it up to pressure.

I was thinking something that was not extended by inflation. I was imagining a simple robot, little more than a cog, that extends the innermost section, clicks it into place (whatever that means) then moves to extend the next section.

It only is intended to hold pressure while fully extended, and it is not required to be reversible. The seal has to be very robust even with flexing. I think there is a risk of harmonics building up in this ring, just like with a suspension bridge. Of course we would take all the steps to minimise that. To grow the station and also reduce flexing, multiple rings could be placed side by side. A square grid cross-section might be a bit tidier than a hexagonal one because as one tube narrows its neighbours could widen.

One tricky bit is that we want the extended cylinder to have a slight curve in it. Maybe we could build a slight curve into every single section, but there might be a way to use sections that are straight cylinders and click into place with a 1 degree angle between each section, say. I have no idea which approach would create the least problems.

btw, a full torus is a reasonable pressure vessel, but what about half or quarter torus? Would the pressure attempt to straighten it?

[speedevil]

btw, a full torus is a reasonable pressure vessel, but what about half or quarter torus? Would the pressure attempt to straighten it?

Yes.
However, you can cancel this with a simple rope between the two ends, adding another as you add a segment and moving the old one over.