A space station is basically a spacecraft with no propulsion. Which boils down to just the habitat module and the payload.
Much like spacecraft, in a science fiction story a space station can become a character all by themselves. Examples include Babylon 5, Deep Space 9, Waystar from Andre Norton's Uncharted Stars, Supra-New York from Heinlein's Farmer in the Sky The Green Hills of Earth and Rocket Jockey, Nowhere Near from Jack Williamson's short story with the same name, Venus Equilateral from the series by George O. Smith, Thunderbird 5 from the Thunderbird TV show, Elysium from the movie of the same name. For an exhaustive list, look up Islands in the Sky: The Space Station Theme in Science Fiction Literature and The Other Side of the Sky: An Annotated Bibliography of Space Stations in Science Fiction, both by Gary Westfahl. The latter has 975 examples.
There are some science fiction stories about star systems that have no colonists living on the surface of planets. All colonists live in swarms of space station habitats. Examples include Downbelow Station by C. J. Cherryl, The Outcasts of Heaven's Belt by Joan Vinge, and the system of Glisten from the Traveller role playing game.
Oh, Werner von Braun had it all figured out in 1952. In six issues of Collier's magazine he laid out a plan to send men to Luna and Mars. First you build a space ferry as a surface to orbit cargo transport (which was the great-grandfather of the Space Shuttle). Then you use it to make a space station.
And it was going to be a beauty of a space station, too. Three decks, 250 feet in diameter, and a crew of fifty. Makes the ISS look like a tin can. This outpost in space was where the Lunar expedition fleet would be constructed.
It would pay for itself as well. Meteorologists could plot the path of storms and predict the weather with unprecedented accuracy. Radio and TV signals could be transmitted all over the globe. Not to mention observing the military activities of hostile nations.
In other words, it would be MacGuffinite.
Why was this marvel never constructed? Because some clown invented the printed circuit. Freed from the tyranny of fragile and short-lived vacuum tubes, technologists could make unmanned satellites for Meteorologists, radio and TV signals, and watching hostile militaries. Such satellites could be assembled and launched at a fraction the cost of a manned station. They also did not require constant resupply missions to keep the crew alive.
If we had followed von Braun's plan, we would have ended up with a fleet of space ferries, a titanic manned space station, a large lunar base, and men on Mars. Instead, we have four overly complicated space shuttles
near the end of their operational life that have been retired, a four man space station due to be de-orbited and destroyed in 2016, and a few bits of space trash on the Lunar surface. And we haven't been back to Luna since 1972. So it goes.
Why doesn't a space station fall down? A station is in "orbit," which is a clever way to constantly fall but never hit the ground. The ground curves away just enough so that the station never strikes it.
Actually, the International Space Station is in a low enough orbit that atmospheric drag decays its orbit. Periodically, Russian resupply rockets have to boost it higher. Otherwise it would de-orbit and burn up in re-entry. Many readers of this website are too young to remember when NASA's Skylab unexpectedly fell.
Many station designs are wheel shaped, and large wheels at that. They are wheel shaped so you can spin them for artificial gravity. They are large wheels so the rotation rate can be kept low enough so the crew does not experience nausea. This is conservatively 3 RPM, though studies suggest some can acclimate themselves to tolerate up to 10 RPM. The Collier space wheel is 250 feet in diameter and spins at 3 RPM, providing about one-third gee of artificial gravity at the rim.
Habitable planets are great! Inhabitants have quaint expressions such as "Free as Air!"
In space, there ain't no free breathing mix. Any breathable air you either brought along or are manufacturing out of local resources. Neither of which is free, or even inexpensive. Air costs money. If you want to breath, you have to pay.
With interplanetary tourists, the "air tax" is included in the fee for their tour package. People living in a space station have to pay their periodic air tax or suffer the consequences. This is why a space habitat is a particularly pure example of a hydralic state. Obey the people who control life support, or you'll find yourself suddenly trying to learn how to breath vacuum.
"Lurker" is a homeless destitute person living on a space station, especially a space habitat. The person figures there are opportunities on the habitat, they spend most of their money traveling to it, when they can find no jobs the money runs out, so they have no money for a space flight ticket to somewhere else. They then move to anyplace they can find in the station, much like terrestrial homeless live under bridges. The space habitat administrators cannot afford to ship the lurkers elsewhere (there are so many of them), so the problem grows. Of course the lurkers are also preyed upon by the criminal underworld.
The term was invented by J. Michael Straczynski for his TV series Babylon 5, more accurately he adapted the term. In internet forums "lurkers" are people who read the forums but do not make posts or otherwise draw attention to themselves. Straczynski noticed this phenomenon when he was discussion the proposed TV show on GEnie, Compuserve, and Usenet back in the early 1990's. He thought the invisible forum lurkers were a good metaphor for the invisible homeless people on the Babylon 5 space station.
Not generally recognized is the matter of Ken Burnside's Three-Generation Rule for space habitats (space stations where people live and raise new generations of children). As Rick Robinson puts it:
In John Varley's Gaea Trilogy, there are some corporations that wait for the 3-gen rule to kill a habitat. Then they swoop in, take possession, vent it to vacuum to sterilize it, and then sell it to some other idealistic group as a unique fixer-upper opportunity.
A more long-term problem is that of Technological Decline. As Joan Vinge pointed out in THE OUTCASTS OF HEAVEN'S BELT: If a planetary colony falls into barbarism, everybody reverts to a non-technological agrarian society.
If an asteroid civilization falls into barbarism, everybody dies.
It takes lots of technology to run the oxygen system, airlocks, spaceships, hydroponics, nuclear reactors, and other items vital for life in space. No technology, no life. In other words, they are a Hydraulic state.
To which I would add:
Naturally a given space station could have several functions.
Mr. Cambias goes on to note that stations can occupy a variety of orbits. Low planetary orbit just above the planet's atmosphere. High planetary orbit at thousands of kilometers. Geosynchronous / Geostationary planetary orbit at an altitude where the orbital period equals one planetary day (useful for communication, observation, powersat, and meteorology). Stellar orbit where the station orbits the local star instead of orbiting a planet. And Trojan orbits where the station occupies a Lagrange point (beloved of L5 colonies)
The size of a station has many terms, none of which are defined. In arbitrary order of size the terms include Beacon (like an interstellar lighthouse), Outpost, Station, Base, and Colony.
The thought occured to some people (most notably Gerard O'Neill) that if the delta-V cost for traveling up and down a planet's gravity well is so expensive, the expense can be avoided if you simply live in space inside a titanic space station. The classic "L5 Colony" was about 32 km long, and held 10,000 inhabitants. Such a colony could earn its keep by harvesting solar energy or with other more shady revenue streams. A quick Google search on "L5 Colony" will reveal a wealth of details. And if you stick an engine on the end, you have a Generation Ship
It sounds very utopian, and it is.
Now, in a Rocketpunk future, when space stations are dotted over the entire solar system (or even the entire galaxy), they might start out as being just a tiny habitat functioning as a Transport Nexus. Yes, they may start as glorified airplane terminals, but they can become more than that. Space stations near research sites can become college towns, ones near mining sites can become mining towns. Then along will come people willing to import and sell things to the inhabitants, and suddenly you've got a city. Think about the TV show Babylon 5, about a space station at the intersection of interstellar transport routes between several star nations. Started as an outer space bus terminal, but grew to become a center of trade and diplomacy.
If the space city has its own revenue stream, it can go even further, and become an independent city state or station-republic. At some point they will be growing fast enough to justify investing in the construction of a full sized L5 colony.
But remember what Thucydides said above about devolution. If the space city's revenue stream dries up, the city becomes a slum, or even a ghost town. Especially if the space city is a boom town, there to supply a fine selection of expensive vices to the local asteroid gold strike or orbit guard military base. If the strike dries up or the base is relocated, the space city will die.
Space habitats appear in science fiction in the Mobile Suit Gundam Wing animes, C. J. Cherryh's Alliance-Union novels, Alexis Gilliland's Rosinante trilogy, George Zebrowski's Macrolife, John Varley's Gaea Trilogy, Sir Arthur C. Clarke's Rendezvous with Rama, and the tv series Babylon 5.
One problem is that you cannot make a small O'Neill cylinder as a pilot project to gain the expertise to build a full size one, due to the nausea caused by the Coriolis effect. Your first one has to be full sized.
A space colony is a particularly pure example of a hydralic state. If citizens make the powers-that-be (defined as "the people who control life support) angry, said citizens will suddenly find themselves trying to breath vacuum. Obey or die. The way to avoid this is with massively redundant life support infrastructure, in an attempt to decentralize control. Of course this only means you do not have to obey the space colony boss, just obey the boss of the segment you live in.
In C.J. Cherryh's Alliance-Union universe, none of the interstellar colonies are actually on an extrasolar planet. Instead they are space habitats in orbit around various lifeless planets (with the exception of Pell). The glaring unanswered question is if you are not going to be using the extrasolar planets, why did you go to the insane expense of using slower-than-light technology to create space habitats in other stellar systems? It would have been about a million times cheaper to just build the habitats somewhere in our own solar system.
If one is colonizing other stellar systems with slower-than-light starships, mass is at a premium. The expense of delta-Ving every microgram up to insterstellar velocities then braking to a halt means you won't be able to carry much of anything. It requires much lower mass to carry the needs for a colony on a human-habitable planet as compared to carrying the industrial machinery required to construct kilometer-long L5 colonies. In fact, such a colony ship might not even carry full grown colonists.
Naturally if you postulate FTL starships, all bets are off. Then it simply becomes a matter of transport costs.
This may or may not boil down to Space Habitats initially being unique to Terra's solar system.
Larry Niven popularized the "asteroid bubble" technique of creating a huge space habitat. Andrew Love notes that if the asteroid is made of stone, once you start to spin it for artificial gravity it will immediately fly into pieces. As he puts it "there are no stone suspension bridges". Stone is heavy and weak, particularly in tension. A 100 meter external radius asteroid made of granite and spun up to 1 gee would put the granite under stresses about twice the expected strength of granite.
You will note that Larry Niven specifies a asteroid composed of nickle-iron.
Many early SF stories fret about the military advantage an armed space station confer upon the owning nation. Heinlein says trying to fight a space station (or orbiting spacecraft) from the ground is akin to a man at the bottom of a well conducting a rock-throwing fight with somebody at the top. One power-crazed dictator with a nuclear bomb armed station could rule the world! Space faring nations would need space scouts for defense.
But most experts nowadays say that turns out not to be the case. A nation can threaten another with nuclear annihilation far more cheaply with a few ICBMs, no station is required. And while ground launching sites can hide in rugged terrain, a space station can hide nowhere. Pretty much the entire facing hemisphere can attack the station with missiles, laser weapons, and propaganda.
Phil Shanton points out that you don't need a huge missile to destroy an orbiting space station, either. In 1979, the U.S. Air Force awarded a contract to the Vought company to develop an anti-satellite missile. It was not a huge missile from a large launch site. It was a relatively small missile launched by an F-15 Eagle interceptor in a zoom-climb. Vought developed the ASM-135 Anti-Satellite Missile (ASAT), and on 13 September 1985 it successfully destroyed the solar observatory satellite "P78-1". This means that an evil-dictator world-dominator nuke-station not only has to worry about every ground launch site, but also every single fighter aircraft.
It has also been modeled that the U.S. Navy could take out a satellite with a Standard Missile 3.
Things are different, of course if the situation is an extraplanetary fleet that remotely bombs the planet to destroy all the infrastructure. The fleet can construct a space superiority platform while the planet is struggling to rebuild its industrial base. Then the platform can bomb any planetary site that is getting too advanced in rebuilding. This is known as "not letting the weeds grow too tall.
James Snead has written a few paper about space infrastructure. Most interesting is Architecting Rapid Growth in Space Logistics Capabilities (PDF file). On page 23 he gives an example of an orbiting space logistics base, including a space dock. Refer to that document for larger versions of the images below.
...the space logistics base’s functions are: (1) housing for travelers and operating crews; (2) emergency care; (3) in-space assembly, maintenance, and repair; and (4) materiel handling and storage.
The example space logistics base consists of four elements. At the top in Fig. 10 is the mission module providing the primary base control facility, emergency medical support, and crew and visitor quarters. The personnel quarters are located inside core propellant tanks that are retained from the SHS used to launch the mission module. The overall length of the mission module and propellant tanks is approximately 76 m (250 ft). Solar arrays and waste heat radiators (shown cut-away in Fig. 10) are mounted on a framework surrounding the mission module to provide additional radiation and micrometeoroid protection.
The second element consists of twin space hangars. These serve as airlocks for receiving spaceplanes and provide a pressurized work bay for conducting on-orbit maintenance of satellites and space platforms.
As shown in Fig. 11, the space hangar consists of a structural cylindrical shell 10 m (33 ft) in diameter, a forward pressure bulkhead containing the primary pressure doors, and an aft spherical work bay. These elements, which define the primary structure, would be manufactured as a single unit and launched as the payload of an SHS. The large, nonpressurized, space debris protection doors would be temporarily mounted inside the hangar for launch and then demounted and installed during the final assembly of the hangar at the LEO construction site. All of the other hangar components would be sized for transport to orbit in the cargo module of the RLVs and then taken through the hangar’s primary pressure doors for installation.
Future logistics supportability is a key feature of this hangar design. The size, weight, location, and access of the internal hangar components enables them to be inspected, repaired, and replaced without affecting the primary structural / pressure integrity of the hangar. With the exception of the space debris protection doors, this would be done inside the hangar when it is pressurized. The ISS-type airlock and space debris protection doors, although mounted externally, would be demounted and brought into the hangar for inspection, maintenance, and repair. For the repair of the primary pressure doors, they would be demounted and taken into the spherical work bay or the other hangar for servicing.
The hangar’s design enables both pressurized and unpressurized hangar operations to be undertaken simultaneously. When the main hangar deck is depressurized to receive cargo or spaceplanes, for example, pressurized maintenance operations would continue inside the 9.8 m (32 ft) diameter spherical work bay and the 2.8 m (9 ft ) diameter x 4.3 m (14 ft ) work compartments arranged along the top of the hangar.
Hangar operations in support of the passenger spaceplanes, as shown in Fig. 12, highlight the improvement in on-orbit logistics support enabled by the large hangars. After entry into and repressurization of the hangar, the passengers would disembark from the spaceplane. Support technicians, working in the hangar’s shirtsleeve environment, would inspect the spaceplane and, in particular, the thermal protection system for any damage to ensure that it is ready for its return to the Earth. While at the space base, the spaceplane would remain in the hangar to protect it from micrometeoroid or space debris damage. Minor repairs to the spaceplane could also be undertaken to ensure flight safety.
The third element is the air storage system. The prominent parts of this system are the large air storage tanks that are the reused core propellant tanks from the two SHS used to launch the twin space hangars. Besides storing air from the hangars, this system also: manages the oxygen, carbon dioxide, and moisture levels; removes toxic gases, vapors, and particulates; and, controls the temperature and circulation of the air within the hangar and its compartments.
The fourth and final element is the space dock. It would be constructed from structural truss segments assembled within the space hangars using components transported to orbit in the RLVs. The space dock would provide the ability to assembly and support large space logistics facilities, such as the space hotels and large manned spacecraft described in the following. It could also used to store materiel and as a mount for additional solar arrays.
The space hangars and space dock would enable traditional logistics operations of maintenance, assembly, and resupply to be routinely conducted in Earth orbit. This is an enabling capability necessary to become spacefaring and achieve mastery of operations in space.
The space logistics base would have approximately 20 personnel assigned. The tour of duty would be 90 days with half of the crew rotating every 45 days. Crew rotation and base resupply would require approximately 32 RLV missions per year per base with 8 spaceplane missions and 24 cargo missions. This would provide approximately 12,000 kg (26,000 lb) of expendables and spares per person per year. At $37M per mission, a ROM estimate of the annual transportation support cost per base would be approximately $1.2B.
While the LEO space logistics base would have sufficient housing capacity to support the 20 assigned personnel and a modest number of transient visitors, it would not be a primary housing facility. Since people cannot simply pitch a tent and “camp out” in space, establishing early permanent housing facilities is an important and enabling element of opening the space frontier to expanded human operations. The architecture of the Shuttle-derived heavy spacelifter and the LEO space logistics base was selected so that the first large space housing complexes, referred to as space hotels, could be constructed using the same space logistics base modules.
A composite illustration of the design, assembly, and deployment of the example space hotel is shown in Fig. 13. This hotel design is configured as a hub and spoke design with a long central hub and opposing sets of spokes attached to the central hub module. This configuration makes it possible to use variants of the space base’s mission modules and space hangars as the primary elements of the space hotel’s design.
Element 1, in Fig. 13, shows the start of the hotel assembly sequence. The central hub module, shown with the SHS’s core propellant tanks still attached, is being positioned at the space logistics base’s space dock. The central hub module would be a version of the mission module used in the space logistics base. Its design would include 12 docking ports around its circumference for attaching the spokes.
Element 2 shows the completed hub and one attached spoke. Two space hangars are located at the ends of the hub and the first spoke is shown attached to the central hub module. In assembling the hub, the core propellant tanks from the two SHS missions used to launch the hangars would be incorporated into the hub to provide additional pressurized volume. This approach would be also used for the spokes. Each spoke would consist of a generalpurpose mission module with the SHS’s core propellant tanks reused for additional pressurized volume. As with the mission module on the space logistics base, the spokes would be surrounded by solar arrays and waste heat radiators. This is what provides their “boxy” appearance.
Element 3 shows the completed 100-person space hotel with two pairs of spokes on opposing sides of the hub. This is the baseline space hotel configuration. Seven SHS missions would be required to launch the hub and spoke modules for the baseline hotel. One additional SHS cargo mission would be used for the solar arrays and waste heat radiators.
This design enables the hotel to be expanded to 6, 8, 10, or 12 spokes. Each spoke would require one additional SHS mission. The 12-spoke configuration would accommodate up to approximately 300 people. Each additional spoke would be tailored to provide a specific capability, such as research and development facilities, tourist quarters, office space, retail space, etc.
Element 4 shows the completed space hotel after being released from the space dock. It also shows how the hotel would rotate about the long axis of the hub to produce modest levels of artificial gravity in the spokes. At about two revolutions per minute, a Mars gravity level is achieved at the ends of the spokes. This use of artificial gravity enables the spokes to be organized into floors (Element 5 in Fig. 13). Each spoke would contain 18 floors with 14 of these available for general use and the remaining 4 floors used for storage and equipment. The spokes would be 8.4 m (27.5 ft) in diameter. This would provide a useful floor area of approximately 42 m2 (450 ft2) per floor. The total available floor area in the baseline configuration would be 2,340 m2 (25,200 ft2). The 12-spoke configuration, having 192 floors total, would have 3 times this floor area—7,026 m2 (75,600 ft2) or about 23 m2 (250 ft2) per person.
An estimate can be made of the number of guests visiting the hotel each year. Assuming a 3:1 ratio of guests to staff, approximately 76 guests would be staying each night in the baseline configuration and 228 guests in the full configuration. With one third of the useful floors configured as guest cabins, two cabins to a floor, each cabin would have a useful area of approximately 21 m2 (225 ft2).* With an average stay of one week, approximately 4,000 guests and 12,000 guests would visit the 4- and 12-spoke hotels each year, respectively.
If each passenger spaceplane carries 10 guests, approximately 400 and 1,200 RLV flights would be required each year. With an additional 25% required for staff transport and resupply, the 4-spoke hotel would require about 10 flights per week and the 12-spoke hotel would require about 30 flights per week. If the RLVs could achieve a one-week turnaround time, and allowing for one in five RLVs being in depot for maintenance, 12 RLVs would be required to support the 4-spoke hotel and 36 RLVs for the 12-spoke hotel.†
At the $37M per flight cost discussed previously for first generation RLVs, the per passenger transportation cost would be approximately $3.7M. With this transportation cost structure, a sustainable space tourism or space business market may not be possible. However, if a second generation RLV could reduce this cost by a factor of 10 to $0.37M per passenger, as an example, then an initial market demand for the baseline hotel may develop and be sustainable. In such case, the annual transportation revenue for the baseline hotel would be $3.7M x 500 = $1.9B and the 12-spoke hotel would be $5.6B.‡ This improvement in transportation costs would also yield a savings of 90%—approximately $1B per year—in the transportation costs to support the LEO space logistics bases. Human space exploration missions would also realize a significant cost reduction.
While developing a conceptual design of a space hotel would appear premature at this early stage of considering the architecture of an initial space logistics infrastructure, several important conclusions emerge that indicate otherwise:
1) Careful selection of the initial space logistics architecture can also establish the industrial capability to build the first space hotels necessary to enable the expansion of human enterprises in space.
2) A commercially successful space hotel will require second generation RLVs to lower further the cost of transportation to orbit.
3) In order for these second generation RLVs to be ready when the first space hotel is completed, the technology research investment would need to begin concurrently with the start of the detailed design of the initial space logistics systems. Conversely, for private investment to seriously consider building the first hotels, significant science and technology progress in developing the second generation RLVs must be demonstrated by the time the initial hotel construction contracts are made.
4) The benefits of reduced space transportation costs will also substantially lower the cost of operation of the initial elements of the space logistics infrastructure, leading to a likely increase in demand for more in-space logistics services.
5) Space hotels and second-generation RLVs may become an important new aerospace product for the American aerospace industry, establishing American leadership in this new and growing field of human astronautical technologies.
6) It is not unrealistic to expect, with the building of an integrated space logistics infrastructure, that hundreds of people could be living and working in space by 2020, growing to thousands of people by 2040 with many of these living in the first permanent orbiting space settlements.
* A standard cabin on the new Queen Mary 2 cruise ship has an area of 18 m2 (194 ft2). A premium cabin has an area of 23 m2 (248 ft2).
† Launch sites for these RLVs would be distributed around the world. This would allow operations at the space hotel to run 24 hours per day since there is no day and night in LEO.
‡ This further reduction could come about through the introduction of a spiral version of the first-generation RLVs where improvements to the high maintenance cost subsystems, e.g., engines, could substantially reduce the recurring costs. Another approach would be development of entirely new RLV configurations—perhaps a single-stage configuration—that would also result in a substantial reduction in recurring costs per passenger through subsystem design improvements and the ability to carry more passengers per trip. A key issue in both approaches is the amortization of the development and production costs. High flight rates, probably dependent on space tourism, would be required to yield an overall transportation cost sufficiently low to enable profitable commercial operations.
This clever design solves the problem of how to quickly assemble a wheel space station, with one tiny little drawback. You see, there is a reason that wheel space stations are shaped like, well, wheels and not like hexagons.
The amount of centrifugal gravity experienced is determined by the distance from the axis of rotation (the greater the distance, the stronger the gravity). So if you want the amount of gravity to be the same, the station has to be a circle.
Now, look at the image below. The segment labeled "SPACE STATION RIGID MODULE" is one of the hexagonal sides. The green lines lead to the axis of rotation (i.e., that is the direction of "up". Note the little dark men figures, they feel like they are standing upright). And the red lines are lines of equal gravity. You will note that they do not align with the module.
In the module, centrifugal gravity will be weakest at the center of the module, and strongest at the ends where it joins with the neighbor modules. Even though the module is straight, the gravity will feel like it is a hill. If you place a marble on the deck in the center, it will roll "downhill" to one of the edges.
As you see, the designers tried to compensate for this by angling the decks, but it really doesn't work very well.
Somewhat more elaborate in conception is the 94 ft wheel-shaped satellite prepared by two design engineers of the Lockheed Missile Division. This celestial laboratory for a crew of 10 weighs 400 tons, and is intended to orbit at a height of 500 miles. Each pre-fabricated section is 10 ft in diameter and 20 ft long, fitted with airlocks, and weighs 10 tons delivered into orbit. Powerful 3-man "astro-tugs" would round up the orbiting packages and couple them up. The entire operation should not take more than a month. The whole design has been investigated in exceptional detail, and is complete with nuclear power reactor and propulsion unit for changing orbit, astronomical telescopes, computer room, space-medical laboratory, airlocks for access, etc. All gravitational worries would be relieved by rotating the vehicle about its hub, which would remain stationary for observational purposes.
The Lockheed space station made an apperance on the 1959 TV show Men into Space. I have not seen this show, but from what I've read it was astonishingly scientifically accurate. Certainly more accurate than most anything from TV or movies in the last couple of decades. Thanks to Drake Grey for bringing this to my attention.
This is from those innocent days before the discovery of nuclear power. The station uses solar power in the form of mercury boilers, since these were also the days before the discovery of the photoelectric effect.
Station desgined by R.A.Smith and H.E.Ross (circa 1940). Again, the station is powered by mercury boilers. The telescope uses a coelostat to counteract the spin of the station. The antenna support arm is de-spun to allow a spacecraft to dock, then is spun up to allow the air-lock module to mate with the station habitat module.
Apparently the artist who did the book cover had seen this old Soviet space station design. I have not been able to discover any details about the station, except that the model is apparently hanging in some Russian museum.