Market research in Japan and America has shown that space tourism has the potential to generate a commercial lammnch market of lOs of launches per day, which is large enough to justify developing a new generation of reusable launch vehicles (1, 2). However, most market research participants would like to stay in orbit for several days or longer (1, 2). Consequently, in order to expand the market, orbiting "hotels" will be needed, enabling guests to enjoy a wide range of entertainments in Earth orbit. As these hotels grow in sophistication, one feature will be spons centers providing guests the chance to enjoy "zero gravity" more actively (3). The design and construction of a simple orbiting gymnasium, which might become the first zero-gravity sports-center, are described below.

The requirement is to provide guests with the possibility of enjoying physical activities in a comfortable zero-gravity room having a volume of several thousand cubic meters. Possible activities will include practising moving in "zero-G", gymnastic exercises, and games and sports for both individuals and teams. Each of these activities will be performed at several different skill levels from beginner to advanced, under supervision of professional staff.

The gymnasium will be attached to an orbiting hotel, from which it will receive services such as electric power, HVAC and station-keeping. The gymnasium will be launched in component form by a cargo version of the " Kankoh-maru" reusable launch vehicle (4), designed as part of the JRS Space Tourism Study Program (5), and will be assembled in orbit by a construction team.

A possible additional function which is also considered briefly is to act as a shelter for all hotel guests and staff in the event of emergencies such as solar-flare radiation or a fire. This would be efficient since the volume of the gymnasium will be sufficient to shelter all residents simultaneously.

The basic design of the gymnasium is a spherical shell 20 meters in diameter, assembled frdm curved pentagonal and hexagonal segments of two sizes (3). 12 regular pentagons and 20 regular hexagons made of aluminium alloy will be joined using flanges and bolts. The inner edges will be seam-welded, and a flexible liner will be attached. The gymnasium will be connected flexibly to the host hotel in order to isolate vibrations.

The gymnasium will contain the following sub-systems.

This must maintain the temperature of the gymnasium air, components and walls within acceptable limits. This will involve both passive design using multi-layer insulation and heaters attached to the external surface, and active systems using cooling water, cold-plates, air-conditioning equipment and temperature sensors.

The atmosphere in the gymnnasium will be monitored and maintained primarily through atmosphere exchange with the central hotel system. That will remove carbon dioxide to the appropriate extent, replenish the oxygen through partial recycling, and remove odours. Air circulation through the gymnasium and the connecting passage to the hotel will be monitored and actively maintained. In order to be used as an emergency shelter, an autonomous carbon dioxide absorption system and reserve supplies of oxygen and nitrogen will be provided.

Electric power for lighting, environmental sensors and control, emergency systems and miscellaneous equipment will be supplied from the hotel's power system, generated from solar panels. If it is to be designed to operate autonomously in the event of emergency, solar panels will also be mounted on the outer surface of the gymnasium, and inter-connected to the power system.

This will be one of the functions of the atmospheric control system. Water will be recovered from the air and recycled as in the rest of the hotel water supply system, to which the gymnasium will be connected.

Fire sensors and fire extinguishing equipment will be designed to suit the special characteristics of fires in zero-gravity. As far as practical, non-inflamnmable materials will be used in the gymnasium fittings.

Telephones will connect the gymnasium to other parts of the hotel. If the gymnasium is to serve also as an emergency shelter, autonomous tele-communication with both the ground and other orbiting vehicles will also be provided.

Movements of the gymnasium relative to the hotel caused by guests colliding with the walls, and movements of the hotel relative to the gymnasium caused by docking vehicles will be damped, using shock-absorbing pistons connected to a flexible metal bellows connection between the gymnasium and the hotel.

Protection will also be required for the gymnasium against orbiting debris. A number of ways of removing orbital debris are being studied, which could become feasible at a time when launch costs are much lower. In the absence of such steps substantial debris shielding will be required, following cumient practice for orbi ti ng facilities.

Lighting, facilities for light meals and drinks_ toilet facilities and an external airlock will also be included. In most respects these will be similar to corresponding equipment in other parts of the hotel. However, in some respects they will need to be appropriate for a gymnasium, such as being protected from impacts. In order to serve as an emergency shelter, there will also be a need for autonomous waste management, to permit continuing use of toilets even in the event that the drainage connection to the hotel is closed.

The basic technologies required for the systems listed above already exist. However, a facility such as a hoteh designed for commnercial customers will require the development of new design codes and standards appropriate for commercial construction operations in orbit. This is an interesting new subject for space engineering design that has yet to be tackled.

From a structural point of view, a sphere is the most mass-efficient shape for a pressurised chamber. Consequently the gymnasium is spherical, assembled from curved pentagonal and hexagonal segments. Using 12 regular pentagons and 20 regular hexagons with equal-length sides, the size of each segment is as shown in Figure 1.

The dominant force acting on such a structure will be the internal pressure of 1 atmosphere causing tension in the walls. As estimated in (3), using a safety factor of 2, the required thicknesses will be about 2.5 mm for aluminium alloy 2219-T87 and 4.2 mm for 6061-T6, giving masses of some 9 tonnes and 16 tonnes respectively. The pentagonal and hexagonal segments are to be joined by bolting flanges along the inside of the edges of each segment, and the inner edges of the flanges will then be seam-welded. For estimation purposes, the total mass is taken conservatively as 30 tons and 50 tons in the two cases.

In addition to the tension in the walls, the structure must resist stresses and fatigue caused by vibration. In order to estimate these, the vibration modes need to be analyzed, and the major frequencies calculated. The vibration modes of a pressurised sphere are a standard case. The vibration modes of the real gymnasium, being a segmnen ted sphere, will be more complex each segment having independent mnodes of oscillation at higher frequency than the structure as a whole. In addition there will be a range of equipment and flttim)gs attached to the structure making the vibration modes non-symmetrical. Finally, the air contained in the gymnasium, with a mass of 4 200 kg, will have additional vibration modes. However, since the overall structure is of relatively low mass, there is ample scope to make it strong enough to withstand these stresses.

The gymnasium will also be subject to mechanical impacts which will cause both structural vibrations and movement of the gymnasium relative to the main body of the hoteL The main sources of these movements will be from guests colIWing with or kicking off from the walls, and from impacts to the hotel dudog docking of ferry vehicles transmitted to the gymnasium through the connection with the hotel main body.

In order to estimate these, we assume that up to 50 people of average mass 80kg use the gymnuim simultaneously. Their movements will be largely random, but if we assume a worst case in which everyone kicks off simultaneously from the walls at 4 m / Sec, some 32 KJ of kinetic energy will be induced in the gymnasium structure. If people cross the gymnasium and impact the opposite wall, this will be repeated after about 2 secona Consequently the gym structure must be able to absorb this amount of energy without damage, and to damp any vibrationt

It is also necessary to prevent motions induced in the gymnasium from being tranmiticci to the hoteL Thus the connection with the hotel is flexible, permitting relative motion of a certain amount Flexibility is achieved by using a flexible metal bellows some 2 metres in diameter to connect the gym to the hotel, and damping by using hydraulic pistons as shock-absorbers. Steel bellows are used today in various construction applications such as large-scale air-conditioning systems.

As assumed above, 32 KJ of kinetic energy may be induced in the gymnasium for a few seconds. By comparison, the US space shuttie approaches the Mir space station with a maximum final velocity of some 0.06 m/sec, impacting with about 200 1 of energy. The 55 ton "Kankob-mam" vehicle docking with a final velocity of even 0.5 In'sec would impart only 7 KJ. In order to damp such vitrafions air-piston snoct-atisorcers will conneci the hotel and gymnasium inside the bellows.

At the launch cost of less than fl million per person that is planned for the Kankoh-maru launch vehicle (4), the cost of crewed activities in Eanh orbit will be much less than they are today. Conacquendy construction activities will be only partially automated, similar to construction operations on Earth today. All the components will be launched in a few payloads of the cargo version of Kankoh-maru and assembled by construction workers either biased inside the host hotel, or working from a dedicated orbital assembly base. Only the initial positioning and bolting of the structmrral segments will require workers to wear space suits and work in vacuum. The main construction operations for the gymnasium are as follows.

Modern tunnelling methods use robots for positioning and fastening massive cylindrical tunnelwall segments around the inside surface of the tunnel. Advanced systems such as Hazama Corporation's "SABIS" (Segment Automatic Building Intelligent System) use position-sensing lasers to control the 3-D positioning, with 6 degrees-of-freedom, of 2500 kg tunnel-segments around the tunnel axis with an accuracy of 0.5 mm (6). SABIS then bolts the segments together automatically. The system adjusts automatically to make allowance for such problems as deformation of the erector structure under variable loading in the Earth's gravitational field.

A smaller, light-weight version of SABIS will accurately position the segments of the orbital gymnasium around the structure's axis, and bolt them together. Each segment will have a mass of only about 500 kg, and effectively no weight. There will also be no problems such as dirt, noise or water ingres. Consequenfly assembly will be much easler than tunnelling operations on Built.

The assembly process will start from the segment attached to the connection with the host hoteL Segments wUl be connected in successive rings as in tunnelling, and the procedure will be practised on Earth in advance. The cargo version of Kankoh-maru will have a cargo bay some S m in diameter, and a payload capacity to LEO of some 5-6 tons. It will therefore be suitable for cai'rying relatively wide objects such as the curved segments of the gymnasium stacked together.

Welding along the inner edges of the flanges will be easier than weldina the outside seams, as the interior will be pressurised and the welding equipment need not be automated. In addition, subsequent inspection and repair will be easier. Welding and inspection could be performed either by crew using specialised equipment or by a specialised robot.

Thermal insulation material similar to that in buildings on Earth with high insulation requirements will be attached to the inside surface of the gymnasium structure. Installation will be straightforward, since the gymnasium will be pressurised. Components of up to about 2 meters in diameter could be brought into the gymnasium from the hotel.

Cabling and ducting for lighting, HVAC and power supplies will be placed between the liner and the interior wall. Their installation will be similar to installation in buildings on Earth, except that manoeuvring large pieces of equipment will be easier due to their weightlessness.

Tools used for such work will be somewhat different from those used on Earth, in ways that have been understood since people lived and worked in the US space station "Skylab" in the early 1970s. For example electric drills will contain a counter-rotating mass; small tools will be connected loosely together to prevent them drifting away and getting lost; and convenient hand-hold and foot-holds will be provided for construction-crew members to brace themselves to exert force in zero-gravity.

The interior wall of the gymnasium will be designed of a material to cushion people's impacts, to be non-inflammable, to be visually attractive, and to be easy to clean. The overall shape need not be smoothly spherical, but may have ribs, for example covering air ducts. It will also have nonprojecting fixing points for attaching sports equipinent such as ropes, nets, goal posts, and so on.

The economnic feasibility of a gymnasium as one facility in an orbital hotel depends on the additional cost of the gymnasium (including both capital and operating costs) being less than the additional price that guests would consider acceptable. This is essentially the same criterion that a company planning a hotel on Earth faces when deciding whether to include a swimming pool.

The cargo version of Kankoh-maru is expected to have a launch cost of ¥1OO million or less for some 5-6 ton nes of payload. Even allowing for some packing inefficiency, launching the parts of the gymnasium should take no more than 10 flights, costing perhaps ¥1,000 million. From the above discussion it is clear that the structure itself will not be difficult to design, manufacture or assemble. In addition the associated systems will mostly be similar to those used in orbiting hotels existing at the time, although the impacts on the design of the host hotel and its sub-systems such as air-conditioning, electric power and drainage will add to the cost. We provisionally assume that the total investment, including launch_ will be some ¥3,O00 million.

In order to be commercially justifiable the gymnasium cost would need to be recovered at perhaps 10% per year if financed in Japan, or perhaps 20% per year if financed in the USA or Europe. Thus the annual amortization cost would be some ¥300-600 million / year, or ¥6 million - ¥12 million / week. If we add 50% to allow for operating and maintenance costs, the required income would be ¥9 million to ¥1 S million per week. Thus if 500 people visited the hotel each week, the cost of the gymnasium would add some ¥l8 000 - ¥36,0O0 per person.

Although many of the cost assumptions made here are only approximate, it seems likely that a cost of this size will be considered acceptable by comparison with a price of some ¥2 million for a short visit to orbit. Consequently, a facility such as this gymnasium would seem likely to be attractive to operators of orbital hotels by providing scope for interesting new entertainments at a relatively small increase in the overall cost of passenger accommodation.

With the publication of the design of the Kankoh-maru passenger launch vehicle (4) and associated proposals for passenger space travel services, the design of space hotels is becoming timely. A gymnasium would provide considerable additional entertainments for guests of space hotels. The above analysis suggests that, at a time when space hotels are being designed and constructed in orbit, design and construction of an orbital gymnasium as an additional facility attached to a hotel would be technically straightforward and economically attractive for hotel operators. More detailed design work on this and other designs of sports center therefore seem worthwhile. An important uncertainty is the mass of orbital debris shielding that will be required, which will depend on the orbital environment at the time of construction.

1. P Collins, Y Iwasaki, H Kanayama and M Ohnuki, 1994, " Potential demand for passenger travel to orbit", Engineering Construction and Operations in Space IV, ASCE, Vol.1 pp 578-86
2. P Collins, R Stockmans and M Maita 1995, "Demand for space tourism in America and Japan, and its implications for future space activities", Proceedings of 6th ISCOPS, AAS in press
3. P Collins, T Fukuoka and T Nishimura, 1994, "Zero-gravity sports centers", Engineering Construction and Operations in Space 4, ASCE, Vol.1 pp 504-13
4. K Isozaki et al, 1994, " Considerations on vehicle design criteria for space tourism", IAF paper no. IAF-94-V.3.535
5. M Nagatomo (ed), 1994, " Space Tourism Special Issue No.2", Journal of Space Technology and Science (Vol.10, No.2)
6. T Matsushita, K Murano, T Sonoda and H Haino, 1989, " Development of 'Segment Automatic Building Intelligent System' (SABIS)", Proceedings of 7th International Symposium on Automated and Robotic Construction ( ISARC)