Rotating Spaceships

[Paul451]

Slightly Longer-term idea here, but what if you followed a sort-of real-estate approach?  Instead of one company building the whole station (expensive), the company builds a "shell" station.  This would have little more than the basics - a Heavy-duty balancing system, a spin-up/down system, and minimal life support.  The company can then sell "land" on the station and make a profit without as much capital expenditure as required to build the entire station.

Effectively, that is a whole station. The minimum build before you can sell "land" is exactly what any commercial space-station would involve.

Power, thermal, lifesupport, pressure shell, MMOD/rad shielding, attitude control, altitude thrusters/OMS, and in the case of a spin station, spin-up/down system and mass distribution control.

[Coastal Ron]

Slightly Longer-term idea here, but what if you followed a sort-of real-estate approach?  Instead of one company building the whole station (expensive), the company builds a "shell" station.

I have a rotating space station design that could be progressively built like that, but I think most of the cost of a rotating space station would be for the station itself - outfitting living spaces would likely be the minority of the total cost.

This would have little more than the basics - a Heavy-duty balancing system, a spin-up/down system, and minimal life support.

Why would there need to be a "spin-up/down system"? That would be a LOT of propellant to do that.

The company can then sell "land" on the station and make a profit without as much capital expenditure as required to build the entire station.

Here on Earth land developers only build out their properties if they have a reasonable expectation that the market is there to buy/lease them. Doing things is in space is so new, that I don't see anyone building a rotating space station unless they have a close to 100% commitment from one or more customers.

[jdon759]

Slightly Longer-term idea here, but what if you followed a sort-of real-estate approach?  Instead of one company building the whole station (expensive), the company builds a "shell" station.  This would have little more than the basics - a Heavy-duty balancing system, a spin-up/down system, and minimal life support.  The company can then sell "land" on the station and make a profit without as much capital expenditure as required to build the entire station.

Effectively, that is a whole station. The minimum build before you can sell "land" is exactly what any commercial space station would involve.

Power, thermal, life-support, pressure shell, MMOD/rad shielding, attitude control, attitude thrusters/OMS, and in the case of a spin station, spin-up/down system and mass distribution control.

&

I have a rotating space station design that could be progressively built like that, but I think most of the cost of a rotating space station would be for the station itself - outfitting living spaces would likely be the minority of the total cost.

Those are good points about it being most of the station already, and it is a problem I am aware of - the most expensive part has to be done anyway. 

But I think you may be overestimating how "complete" my suggested station is.  I do not expect the initial "shell station" to have much more habitable volume than a reentry capsule. 
Modules with their own life support, pressure shell and MMOD/rad shielding would be the "buildings" that are built on the "land" of the shell-station's structure.  Power and stationkeeping would be all that is provided to these buildings by the shell-station.  This requires a heavy-duty balancing system because the station is assured to be imbalanced due to decentralised design and non-simultaneous construction. 
It's analogous to a Stanford torus where the dwellings within the torus are not built at the time the torus is "complete", but are added later by the people living there.

[Coastal Ron]

Those are good points about it being most of the station already, and it is a problem I am aware of - the most expensive part has to be done anyway.

Who is going to pay for that? What is their return on investment (ROI) projection for building a station with no paying customers?

We space enthusiasts LOVE to solve hardware problems, but in the real world it is money that determines the solutions.

But I think you may be overestimating how "complete" my suggested station is.

Of course we are, since only you know what your station looks like, and how it operates . And you don't need to show us, since we're just having a friendly conversation here.

I do not expect the initial "shell station" to have much more habitable volume than a reentry capsule. 
Modules with their own life support, pressure shell and MMOD/rad shielding would be the "buildings" that are built on the "land" of the shell-station's structure.

Modular construction is something that is done here are Earth, which makes it easier to understand the many different approaches that can be done. What you are suggesting though may not be as easy as you hope, since the advantage of building a fully functional structure is that it is easier to finish for a customer.

What you seem to be suggesting is something akin to an RV park, where everyone is responsible for just about everything - including air and water apparently.

It is certainly one approach - who knows, maybe even a good one. But as I mentioned earlier what determines the solutions we end up using in the real world is the money that those funding something THINK they can make in return, and what potential customers are willing to pay. So I would ask that you consider what everyone would be willing to pay for.

For instance, with the modular approach you propose the upfront cost to a user/customer is very high, as is the operating cost. Plus there is little to no redundancy between customer modules. But if you lower the cost to occupancy, maybe you would get more users/customers? Just as in engineering you have do tradeoffs, so it is in the world of money too.

My $0.02

[mikelepage]

But I think you may be overestimating how "complete" my suggested station is.  I do not expect the initial "shell station" to have much more habitable volume than a reentry capsule. 
Modules with their own life support, pressure shell and MMOD/rad shielding would be the "buildings" that are built on the "land" of the shell-station's structure.  Power and stationkeeping would be all that is provided to these buildings by the shell-station.  This requires a heavy-duty balancing system because the station is assured to be imbalanced due to decentralised design and non-simultaneous construction. 
It's analogous to a Stanford torus where the dwellings within the torus are not built at the time the torus is "complete", but are added later by the people living there.

I do think this decentralised format is a good way to approach the problem, but the real-estate metaphor implies things maybe you didn't mean. This isn't an apartment or house you're offering. If I'm reading you right, you're literally suggesting building a framework that functions as the equivalent of a new green-fields housing development site that sells empty plots of land, with standardised access to basic utilities.  That land has inherent value based on its location, (as would the spots for spin-G payloads on a spin-G structure). You probably wouldn't even be "selling" that land, so much as renting it.

Personally, I would not even be including any kind of pressurised volume, or need to include significant balancing devices since you'd just put in the spec that customers have to produce two payloads that balance each other (or maybe there would be a secondary market where smaller customers are matched together to achieve this). What you're selling is access to a standard amount of partial-gravity through spin G, and since you've got to build, assemble and power the structure anyway, you can include a power supply as part of your service.

I think it's especially valuable to delineate between the customer payload and structure since, at least for the early spin-G experiments, there's probably extra value in the ability to return the entire customer payload back to Earth for analysis.

[Paul451]

The problem I have with the real-estate model is that with an "empty" site on Earth, there's common "infrastructure" that we take for granted that doesn't exist in space. Air. (Duh.) But also everything that it allows, such easier thermal control (just circulate air through your building, or through your heat-exchanger,) and easier access to neighbouring sites and access to other infrastructure like roads. Seemingly little things like pressure automatically equalising in any system that isn't gas-tight. For example, in sewerage systems. Makes construction and use vastly easier.

Similarly, infrastructure is one-way, any "loop" is well outside the greenfield-site (possibly even environmental, like the hydrological cycle.) This allows things like fresh-water-in and waste-water-out to be treated completely independently. Therefore one can be pressurised and water-tight (although not very water-tight in practice, because it all ends up back in the rivers/seas/air or groundwater somehow), the other can be at ambient external pressure through the entire system with correspondingly lower construction/maintenance costs. (Indeed, the trick is to add vents.)

In space: You only have radiative cooling. Heating (both internal and external) is more complex than on the ground. Life-support has to be handled by each module, or connected through a common shared infrastructure which is two-way. Each module has to be gas-tight, which means anything passing in or out has to go through gas-tight (not just water-tight) lines/pipes. (And to stay gas-tight, even if something goes wrong elsewhere.) But importantly, any movement of fluids/gas in and out of the module has to be balanced to ensure you aren't messing up the internal pressure. If you bust a pipe in your home, it makes a mess. If you bust a pipe in a space-station, it might allow your air to escape back up the pipe and into space. You might die.

It's not just that "it's harder". It's "different" to the point that not only does the analogy break down, but that such a method of construction would end up costing everyone more than simply not doing it. It wouldn't even save money in the development stage of the core station (pushing costs forward to the clients, reducing buy-in for early investors.)

A closer model is the apartment/office building (and perhaps even more limiting.) Clients "furnish" their rented/leased volume with whatever specific equipment they need, but everything else is part of the "building".


That said, I agree that the "mass balancing" requirement is probably overblown. This isn't a wheel-on-an-axle, where even slight imbalances cause vibration, free-rotating objects find their own centre-of-rotation. Beyond the intermediate-axis issue discussed {waves hand} over a bunch of threads, the system will be innately stable. The only requirement for a fixed centre-of-rotation is for docking. And IMO it's better to design a docking system that can cope with off-axis approach, rather than require the station (and approaching spacecraft) to be perfectly balanced around a (shared) fixed physical axis. It's yet-another-system-to-develop (yay!) but once solved, it eliminates so much hassle.

[Coastal Ron]

The problem I have with the real-estate model is that with an "empty" site on Earth, there's common "infrastructure" that we take for granted that doesn't exist in space. Air. (Duh.) But also everything that it allows, such easier thermal control (just circulate air through your building, or through your heat-exchanger,) and easier access to neighbouring sites and access to other infrastructure like roads. Seemingly little things like pressure automatically equalising in any system that isn't gas-tight. For example, in sewerage systems. Makes construction and use vastly easier.

I've worked on a design for a Earth-level gravity rotating space station that would have provided open space for future tenants, but the infrastructure issue, as you point out, becomes a big challenge. Bottom line is that there were too many challenges to solve with that design, and it was highly unlikely that anyone would need a 1G station anytime soon. Instead I'm focused on a Mars-level gravity rotating space station that will be fully built out, though some internal spaces could be customized, and I'm using the same approach for my rotating spaceship designs.

That said, I agree that the "mass balancing" requirement is probably overblown. This isn't a wheel-on-an-axle, where even slight imbalances cause vibration, free-rotating objects find their own centre-of-rotation. Beyond the intermediate-axis issue discussed {waves hand} over a bunch of threads, the system will be innately stable. The only requirement for a fixed centre-of-rotation is for docking. And IMO it's better to design a docking system that can cope with off-axis approach, rather than require the station (and approaching spacecraft) to be perfectly balanced around a (shared) fixed physical axis. It's yet-another-system-to-develop (yay!) but once solved, it eliminates so much hassle.

Agreed.

[mikelepage]

I took my time thinking about how to reply to this thread, since I fell into my own trap with the post above, with the real estate-model implying things I didn't mean, and Paul you assumed I was arguing a point that I wasn't arguing. Of course there is "infrastructure" present on Earth not available in space in the form of innumerable physical and environmental services. It's this leap bolded below that I disagree with.

It's not just that "it's harder". It's "different" to the point that not only does the analogy break down, but that such a method of construction would end up costing everyone more than simply not doing it. It wouldn't even save money in the development stage of the core station (pushing costs forward to the clients, reducing buy-in for early investors.)

I can agree with everything you said before that, and agree that the overall cost will be increased, but still argue that a division of function/outsourcing of components can make the whole process more cost-effective for everyone involved - and that this is a better way to approach the problem. The entities that want space on a rotating spacecraft don't necessarily want to build the spacecraft itself, but they will find value in being able to build the module/habitat that is to be placed on such a spacecraft.

Perhaps a better analogy is of launching a cubesat aboard a ride-share service (which also has a higher overall cost than if one company built the launcher and all the cubesats). It's an example of a situation where one fully functional spacecraft provides a service to a set of other (fully-functional) spacecraft. The advantage is not in the overall cost, but in making it cost-effective for a wide range of different entities interested in space for different reasons. It's because of the modular nature of cubesats that space has become accessible to small companies, universities and even schools, by reducing the significant amount of overhead involved in actually getting to space. And the rideshare business model has enabled a new class of providers that wouldn't see any particular benefit in launching a spacecraft for a particular university science project, but are happy to take money from people who do.

I think it makes sense for a spin g provider to use the same business model. I suppose you would use the same business model even if you were just offering dedicated spaces within a single, fully integrated spin g spacecraft, but then the customers don't get anywhere near as much ability to customise their setups, and the upfront cost is much larger (upfront cost being the reason that has sunk every previous spin g spacecraft proposal).

In a "spartan" spin g provider scenario, you have the provider company simply has to focus on finding a cost effective way to build a framework with the ability to produce spin gravity (a difficult enough task all by itself), and then arrange the launch of separate customer modules, which are standardised to a spec in the same way that cubesats are. This would also mean you can have the modules swapped in and out of the framework as required/desired, rather than having to deorbit a whole space station when only parts of it are reaching their end-of-life dates. Yes, you get a lot of duplication of function/cost, but that is spread among the customers and occurs later in the development process, and it also means the resulting craft has significant redundancy, which is not a terrible thing.

Maybe the best permutation of of the real-estate analogy, is to compare this to a caravan or trailer park, where basic services are provided to vehicles which are functional in other settings beside this one. I know "trailer park" has a specific connotation in the US, but here in Australia I'm thinking of the places where "grey nomads" on holiday will set up camp for a few days to a few weeks before moving on, at a much reduced cost as compared to staying in a dedicated motel or hotel.

[Paul451]

and Paul you assumed I was arguing a point that I wasn't arguing.

I was more replying to the originator of real-estate metaphor, jdon, and the general concept.

such a method of construction would end up costing everyone more than simply not doing it. It wouldn't even save money in the development stage of the core station

I [....] still argue that a division of function/outsourcing of components can make the whole process more cost-effective for everyone involved - and that this is a better way to approach the problem. The entities that want space on a rotating spacecraft don't necessarily want to build the spacecraft itself, but they will find value in being able to build the module/habitat that is to be placed on such a spacecraft.

I'm not sure I can explain this clearly, so apologies in advance:

I don't think you can actually divide anything in a way that saves money even for just the developer of the core alone. The division itself makes the core a harder problem to solve than a "regular" space-station. The developer of the core will still have the burden of, in effect, designing the whole spin-station and all systems, but also of designing a way to divide those systems out to customer such that they can't detrimentally affect the core (or vice versa) nor each other's modules via the shared infrastructure. And also designing the standard for hot-swappable habitat-modules.

If adding modules to a spin-station were just a matter of plugging in power, sure, why not. It's when you factor everything required, not only for supporting humans, but for supporting the systems required to support humans.

Hopefully one day, down the road, when all this spin-g & space-station stuff is already off-the-shelf and technologically mature and boring and cheap, then there might be a business model for these kinds of variants and innovations.

But jdon's suggestion was that this would make things easier for an early station. Maybe even the very first spin-g station. That's where I think it only makes things harder.

[Twark_Main]

As for the cubesat analogy....


Modularity helps, but cubesats are cheap mainly because they're small. Space station modules are not small.

Space station modules also kill people when they fail, so all the additional QA and testing adds a fair bit of cost. If a cubesat fails, nobody dies.

I think we'll see nano-racks inside a rotating space station, allowing universities to fly cheap payloads. But I doubt we'll see university-level budgets (or even single company-level budgets) launching entire habitation modules any time soon.


I think the station-to-module connection is the wrong place to put the customer interface. To borrow a phrase, "rotating space stations are not LEGO." Just like rockets, the whole thing really needs to be designed all together (structurally, infrastructurally, crew circulation and emergency egress, rotational balance, etc). This is less true for purely microgravity stations like the ISS, of course.

[Slarty1080]

The problem I have with the real-estate model is that with an "empty" site on Earth, there's common "infrastructure" that we take for granted that doesn't exist in space. Air. (Duh.) But also everything that it allows, such easier thermal control (just circulate air through your building, or through your heat-exchanger,) and easier access to neighbouring sites and access to other infrastructure like roads. Seemingly little things like pressure automatically equalising in any system that isn't gas-tight. For example, in sewerage systems. Makes construction and use vastly easier.

I've worked on a design for a Earth-level gravity rotating space station that would have provided open space for future tenants, but the infrastructure issue, as you point out, becomes a big challenge. Bottom line is that there were too many challenges to solve with that design, and it was highly unlikely that anyone would need a 1G station anytime soon. Instead I'm focused on a Mars-level gravity rotating space station that will be fully built out, though some internal spaces could be customized, and I'm using the same approach for my rotating spaceship designs.

That said, I agree that the "mass balancing" requirement is probably overblown. This isn't a wheel-on-an-axle, where even slight imbalances cause vibration, free-rotating objects find their own centre-of-rotation. Beyond the intermediate-axis issue discussed {waves hand} over a bunch of threads, the system will be innately stable. The only requirement for a fixed centre-of-rotation is for docking. And IMO it's better to design a docking system that can cope with off-axis approach, rather than require the station (and approaching spacecraft) to be perfectly balanced around a (shared) fixed physical axis. It's yet-another-system-to-develop (yay!) but once solved, it eliminates so much hassle.

Agreed.

I'm sure that issues with rotational stability are solvable, but I wouldn't be too quick to dismiss these problems. Rotating structures often behave in very counter intuitive ways:

[mikelepage]

That said, I agree that the "mass balancing" requirement is probably overblown. This isn't a wheel-on-an-axle, where even slight imbalances cause vibration, free-rotating objects find their own centre-of-rotation. Beyond the intermediate-axis issue discussed {waves hand} over a bunch of threads, the system will be innately stable. The only requirement for a fixed centre-of-rotation is for docking. And IMO it's better to design a docking system that can cope with off-axis approach, rather than require the station (and approaching spacecraft) to be perfectly balanced around a (shared) fixed physical axis. It's yet-another-system-to-develop (yay!) but once solved, it eliminates so much hassle.

Agreed.

I'm sure that issues with rotational stability are solvable, but I wouldn't be too quick to dismiss these problems. Rotating structures often behave in very counter intuitive ways:

Yep, Slarty that's the video I often link to when talking about what the intermediate axis problem.

Also Ron, I see you mentioned the RV park analogy above my post where I said it was like a trailer park - not trying to pass off your idea as mine - just didn't see it before.

As for the cubesat analogy....

Modularity helps, but cubesats are cheap mainly because they're small. Space station modules are not small.

Space station modules also kill people when they fail, so all the additional QA and testing adds a fair bit of cost. If a cubesat fails, nobody dies.

I think we'll see nano-racks inside a rotating space station, allowing universities to fly cheap payloads. But I doubt we'll see university-level budgets (or even single company-level budgets) launching entire habitation modules any time soon.

I think the station-to-module connection is the wrong place to put the customer interface. To borrow a phrase, "rotating space stations are not LEGO." Just like rockets, the whole thing really needs to be designed all together (structurally, infrastructurally, crew circulation and emergency egress, rotational balance, etc). This is less true for purely microgravity stations like the ISS, of course.

I should clarify that my thinking on this is mainly framed by my aim to implement a small ISS-deployed spin-G free-flyer, carrying payload modules around the size of 6U (long) cubesats. Treading a bit of a fine line here because of IP concerns, but basically we're designing something where the (deployable) framework can be taken up in a Dragon2 trunk, while the pressurised modules go up (and come back) in the pressurised section of Dragon. Also working on a conops that occurs at the beginning and end of long duration crew missions, where the framework and modules can be manipulated by the station's robotic arms, so astronauts inside ISS would operate the Kibo airlock/robotic arm (JEM-RMS) and Canadarm2 in tandem to install/uninstall modules from the framework before setting it free/after grappling it.

Agreed you have to design the whole thing at the beginning, but making the modules interchangeable like cartridges would be highly desirable if it can be done, because you can then run iterative 3-6 month partial g experiments, reusing the framework component multiple times. Being able to bring those modules down from orbit after completion would be invaluable for a number of fields, so I'm pretty sure there's a saleable service there - if we can get it to be interchangeable.

That said, I also agree it becomes far more complicated once you scale up to human-sized spin-G spacecraft, and many elements obviously would not scale up. Cross that bridge when we come to it.
 

[Paul451]

I should clarify that my thinking on this is mainly framed by my aim to implement a small ISS-deployed spin-G free-flyer, carrying payload modules around the size of 6U (long) cubesats. Treading a bit of a fine line here because of IP concerns, but basically we're designing something where the (deployable) framework can be taken up in a Dragon2 trunk, while the pressurised modules go up (and come back) in the pressurised section of Dragon.
Agreed you have to design the whole thing at the beginning, but making the modules interchangeable like cartridges would be highly desirable if it can be done, because you can then run iterative 3-6 month partial g experiments, reusing the framework component multiple times. Being able to bring those modules down from orbit after completion would be invaluable for a number of fields, so I'm pretty sure there's a saleable service there - if we can get it to be interchangeable.

Deeply cool.

Most (perhaps all?) of my objections don't apply here. The modules only need power/data from the core/frame, and maybe common comms via the data-bus. (And the attachment system itself.) Limiting clients to explicit cubesat standards eliminates most of the other complexities with client-developed modules. No issues in docking while rotating (or having to despin for every dock/undock, which isn't viable for a multi-client manned station) and therefore no real balance issues, greatly simplifying the core. (Beyond the known things like the intermediate axis problem, which are solved in design, not ops.) So even moving masses, like animals or fluids/reactants in the client-modules, won't be an issue.

I genuinely hope it works.

(And not just for your sake, although that too. I think it's the kind of thing that we need more of, along with the increase in commercialisation we're seeing already. Another step making doing things in space just an extension of doing something on that scale on Earth, not a whole separate kind of activity.)