current revision: 11 July 2000
first publication: 4 August 1998

Abstract:
This document describes a conceptual design for a transportation system which could be built to carry passengers into space for tourism. It's futuristic only in that it has not been built yet. It does not rely on warp drive or any other magic technology, but rather adapts concepts developed or proposed for NASA activities.

The system is based on a "space elevator" and space hotel in low Earth orbit. The hotel would orbit 775 miles above the Earth, and would suspend a space dock 160 miles above the Earth, via a hanging tether. Passengers and cargo would be brought to the dock by a new suborbital reusable launch vehicle, and would travel up the tether via a space elevator. The launch vehicle latches onto the dock, and is carried back to the launch site.  The dock moves at only 79% of orbital velocity, which quadruples the payload capacity of the launch vehicle.

An upward deployment tether can be extended from the top station to release or capture payloads onto or off of a trajectories to higher orbits or to the Moon.  The tether system can reuse orbital momentum taken from returning crafts, so that for balanced round trip travel, no propellant is consumed (once the payload reaches the dock).

Motivation

Enabling Technology

The Elevator System

Launch Sites and Schedules

Construction

Cost per Passenger

Beyond Low Earth Orbit

Hazards and Safety Concerns

Radiation Concerns

Conclusion

References, Notes, & Website Links

Back to Site Index Page

Motivation:
Many people want to go into space. Most NASA and aerospace industry insiders probably believe that "someday" tourists will be able to travel into space and visit space hotels. But the general assumptions is that to get from the current situation, in which the only commercial use of space is for unmanned communication satellites, to the space tourism era, some other space industry (perhaps manufacture of exotic drugs or airplane parts, production of electricity for use on Earth, or mining for important minerals that are scarce on Earth) must blaze the trail first.

However, no such industry has emerged. In fact, tourism is about the only industry in which a space based product/service does not have to match the cost of the terrestrial equivalent. Of course, a higher cost will imply a smaller customer base, but the first few space hotels won't require millions of customers, only tens of thousands. And as is the case with any high tech product or service, the cost will come down with passing time and increasing volume (thus "opening the frontier" for other industries).

If one ignores the desire for a bridge industry, and asks what sort of transportation system could be built to serve a space hotel, the space elevator is a clear choice. The unfamiliarity of the space elevator apparently results from its lack of terrestrial applications (whereas rockets can be used for weapons and fireworks), its new level of feasibility (the one described here uses a graphite fiber that did not exist in the Apollo era), and the fact that it is better suited to the frequent, identical missions of tourism than the infrequent one-of-a-kind missions common in the early space program.

Science fiction authors have only occasionally explored the utility of space tethers (e.g. an episode of the T.V. show Star Trek Voyager and a few of books by authors such as Arthur C. Clarke).

Enabling Technology: RLVs and Space Elevators:
It has long been recognized that only a fully Reusable Launch Vehicle (RLV) could make space travel truly affordable. And in recent years, there has been considerable government interest in developing a launch vehicle with Single-Stage-To-Orbit capability (i.e. no booster stages are discarded during flight^note_20).

However, for round trip travel (e.g. tourism), space tethers used in conjunction with sub-orbital RLVs, provide a much better approach. Tethers (essentially long ropes connecting space craft or other bodies) allow momentum and energy transfer, so that energy removed from descending space crafts can be transferred to ascending crafts. The ultimate realization of the concept is the "synchronous space elevator", wherein a gondola travels along a tether stretched between the surface of the Earth and a counterbalance beyond geosynchronous orbit, 22,300 miles away (this idea was made famous by Arthur C. Clarke, but the invention as a cable is credited to Y.N. Artsutanov in 1960 and as a tower to Tsilkovsky around 1900) ^note_1.  Not surprisingly, the full Earth-to-orbit synchronous space elevator is not yet technically feasible (due to material strength limitations).

However, a shorter version of the space elevator, up to about 1,000 miles long, is feasible with existing tether materials, and is an extremely powerful way of making low Earth orbit more accessible (this is a newer idea that seems to have originated with Hans Moravec in the mid 1970's). Space elevators (and hanging tethers in general) take advantage of the fact that higher altitude orbits have lower velocity required for orbit. One end hangs below the Center Of Gravity (COG) of the system, and therefore is traveling below orbital velocity and feels downward gravity. The other end hangs above the COG, travels faster than orbital velocity, and therefore feels the upward pull of centrifugal acceleration.

For the case of a space hotel, the space elevator (say 860 miles long in this case) links a zero gravity station orbiting 775 miles above the Earth, with a space dock hanging only 160 miles from the surface, and a low gravity (low "Gee") station 1,020 miles above the Earth. The low Gee station feels about 1/10 of Earth normal gravity except it's directed upwards, and serves to counterbalance the force of the hanging dock (the Moon-like gravity makes it an interesting place for hotels rooms). The dock is much easier to reach by rocket than a zero-Gee station would be, since it travels at Mach 19.5, only 79% the full Mach 25 required for orbit (more precisely: 13,800 m.p.h. vs. 17,450 m.p.h.).

This 21% reduction in required launcher velocity results in approximately a 300% increase in payload capacity, for a given vehicle launch mass. This large increase is due to the fact that for a single staged vehicle burning oxygen and hydrogen (state of the art), for a Mach 25 flight, the fuel must be 88% of the launch mass, so typical design goals for this type of craft call for a payload of only 1.5% of the launch mass (fuel percentage requirement assumes an unpowered re-entry and landing). Whereas for the Mach 19.5 flight, the fuel need only be 82% ^{note 2}, so the payload can be a much larger percentage of the gross launch mass (for comparison, with the Space Shuttle, the hydrogen/oxygen fuel is 85% of the gross, not including the solid rocket boosters). Significantly, some of the extra breathing room in the mass budget can be dedicated to optimizing for reliability and low maintenance rather than minimum mass.

The precise payload increase using the space elevator will depend on the launch vehicle. However, the increase in mass in orbit (i.e. the payload plus empty rocket mass) for a given lift off gross mass, will always be 44% for hydrogen-oxygen engines (including scramjets with rocket augmentation) and 68% for kerosene-oxygen engines. This percentage applies to the payload plus empty rocket mass so the gain is much larger for single stage rockets (since most of the mass in orbit will be empty rocket) than for the small upper stage of a multistaged rocket.

For the RLV development, this velocity reduction could be the key to viability. NASA's X-33 program ^{note 22} is intended to mature and demonstrate technologies necessary for a follow-on full scale single stage to orbit vehicle, with the hope that a private company will then be willing to fund the follow-on development. However, the biggest technical uncertainty is the feasibility of the mass budget, which the X-33 will fall far short of demonstrating. The use of the space elevator allows the launch vehicle's dry weight be be about 50% heavier (for the same gross weight and payload), thus greatly reducing investment risk.

An added advantage of the lower required launcher velocity is for an aborted launch, the launch vehicle will never land more than about 2000 miles down range (farther for slow accelerating scramjets). This on-continent abort eliminates the need to provide a viable water landing procedure for emergencies.

When the launch vehicle arrives at the space dock, it latches on, and will hang from the dock (feeling about 35% of Earth normal gravity), until it is ready to return to Earth (presumably, the next time the dock passes over the launch site). The passengers and/or payload will move from the launch vehicle into one or more gondolas that will move along the length of the tether, up to the hotel stations. The gravity felt by the passengers will gradually decrease during the ride up (which could be many hours long, potentially helping to prevent space sickness), becoming negative beyond the zero-gravity point, so that at the counter balance station, one could look "upward" and see the Earth.

This type of tether system would be much too massive to be practical for the monthly mission rates required for government and scientific purposes (though there are other types that do not use gondola track and therefore can be scaled down for weekly missions). But it's perfectly reasonable for the daily flight rates required for tourism. In fact, the mass of tether/gondola system (i.e. space elevator) is of the same order of magnitude as the hotel itself (and two order of magnitude greater than the mass of the launch vehicle, which allows the system to remain in a relatively round orbit both before and after the launch vehicle docks).

The Elevator System:
The space elevator system consists of the dock, the hanging tether, the gondolas, the gondola track running the length of the tether, the solar electric power system, and the momentum (& drag compensation) engines.


The Dock:
The dock hangs at the lowermost end of the tether. It has parking spaces for three or four gondolas, berths for two or three launch vehicles, a pressurized & climate controlled passageway connecting the vehicles to the gondolas, and cargo loading equipment.

Each berth has a winch which lowers a magnetic docking grapple to the capture/release altitude of 90 miles; about 70 miles beneath the dock (this makes the effective length of the elevator system 70 miles longer for very little extra cost and weight, but still allows a launch vehicle departure a single orbit after an arrival, assuming the winches can move the vehicles at over 75 m.p.h.). The separation also allows the launch vehicle to fire rocket engines during the docking maneuver without damaging the dock or vehicles parked there.

If less frequent dockings suffice (e.g. for an equatorial orbit), then the dock altitude can be raised to somewhere around 300 miles (limited by the start of the radiation belt). This reduces the total mass of the elevator system by around 25%, at the cost of a bigger winch and longer tow from the docking site to the dock.

Prior to docking, the launch vehicle will extend a steel docking plate that will latch onto the magnetic docking grapple upon contact. The grapple will have rendezvous radar, differential GPS (the Global Positioning System), and probably small rocket thrusters to stabilize its position.

Note that although the sketch shows a Delta-Clipper inspired launch vehicle (vertical takeoff and landing), any type would work, including VentureStar, Roton, or a Horizontal TakeOff and Landing scramjet ^{note 23}.  It might be helpful however, to be able to fire the engines downward during docking so that it's easier to correct position errors during approach.

The Hanging Tether:
The tether is the super strong structural element that holds its own enormous weight, plus that of the dock, the gondola track, and the counter balance stations. It also must withstand the intense radiation of the Van Allen belt. To make the tether as long as possible for a given mass budget (which makes the dock move as slowly as possible and therefore boosts launcher payload capacity), it's necessary to have the strongest material. In particular the figure of merit for tether material is the tensile strength divided by the density.

Available materials are improving all the time, but rather than extrapolate what might be available in a few years, the design described here uses existing materials. The length of the tether (which results in the Mach 19.5 dock speed) was selected to be about a large as could be built with two flights per day for a year (to a similar elevator system). Alternatively, this one year construction criterion would suggest a Mach 21 tether for a five flights per week rate, and Mach 18 with five flights per day.

The material with the best figure of merit is said to be Spectra-2000, which is a polyester fiber ³. It's design tensile strength is about 1.35 GPa (the ultimate strength is 3.25 GPa), and it's density is 0.97 g/cc, so it's figure of merit is about 8 times better than that of steel (1.76 GPa and 7.85 g/cc including design margin), and the feasible tether's length is squareroot(8) times longer. The problem with Spectra (and plastic fibers in general) is that it doesn't last long in space at the altitudes of interest due to the radiation of the Van Allen belt.

Carbon fibers (such as used for graphite-epoxy tennis rackets, etc.) are radiation resistant ⁴ and have a figure of merit very close to that of Spectra-2000.  Thornel T-40 from Amoco Performance Products ⁵ has an ultimate tensile strength of 5.65 GPa and a density of 1.81g/cc (with a 2.4:1 design margin, the tensile strength is assumed to be 2.35 GPa).

The following graph shows the tether's thickness profile (versus altitude), bases on construction with Thornel T-40 Carbon fibers (although Spectra-2000 could be used for the portion below a few hundred miles altitude). The tether will be built as several parallel cables, but the graph gives the diameter of a single cable with the same cross sectional area.

The tether mass was calculated to be 34 million pounds ⁶ (for the tether, dock, and solar panels) assuming: the loaded dock mass is 1.25 million pounds (including two launch vehicles and three gondolas), the gondola track and electrical cabling mass is 1.34 lbs/foot (which comes to 6 million pounds), the solar panels are 4.2 MegaWatts at 100 tons/MegaWatt, and the counterbalance station mass is 22 million pounds (to bring the effective gravity from 1/10 to a full lunar 1/6 Gee, the tether would be extended to 38 Mlbs and the counterbalance mass decreased to 8.6 Mlbs at 1190 miles). The density of the tether material was padded 20% to allow for stiffening trusses and cable jacketing (to reduce erosion by atomic oxygen); and the material strength was decreased by a factor of 2.4 for safety margin (and redundancy).

Because of the risk of a debris impacts (space dust, micrometeoroids, paint flecks from old satellites, etc.) which will frequently occur, it's desirable to make the tether as several parallel cables. One possible configuration would arrange the cables such that the cross section formed a three pointed star, several feet in diameter. Three sets of gondola track would be at the center of the star. This arrangement allows the cables to be widely separated to minimize the likelihood that a single small meteoroid would sever multiple cables. It also allows the gondola tracks to be rigidly mounted to one another, to help insure that gondolas on adjacent tracks are properly separated. The arms of the star would not be completely independent cables, but rather cables interconnected as Hoytapes (a flat Hoytether), a net-like configuration which maintains adequate strength even with many single-cable cuts, at the expense of about a two to one redundancy of material (the Hoytether was developed by Tethers Unlimited, see link to their site ⁷).

Rigid trusses spaced along the tether (maybe only one per mile to minimize total truss mass) would hold all the elements in position, and redistribute the load to remaining cables should one or two be severed (this also makes it easier to replace damaged sections of the tether). Much lighter cable ties might be spaced in between the main trusses, to transfer lateral coriolis loads from the gondola track to the main tether cables (the coriolis acceleration happens because the gondola gains tangential velocity as it moves up the tether).

The epoxy that is used to make graphite composites isn't radiation tolerant ⁸, so it can't be used to make the trusses, however, aluminum should be fine.

It should be noted that carbon fibers are electrically conductive. This means they must be insulated to protect them from the sparse plasma that fills space in low Earth orbit (since the Earth's magnetic field will create an electric potential across them). On the other hand, the tether could also be used as a neutral conductor in the electrical system, and therefore add a bit more redundancy.

Other Tether Types:
It has been observed by tether pioneer Hans Moravec and others that a rotating tether would have a lower mass for a given dock velocity than the hanging tether of the type described here.  This concept would certainly be useful for cargo flights (particularly to destinations beyond low Earth orbit, because in this case the tether need not be climbed and therefore no gondola track is required), but would be inconvenient for several reasons for hotel applications in low Earth orbit. The rotation might be annoying for hotel guest. A rotating tether would also dip the launch vehicle in the radiation belts every rotation (if the tether reaches an altitude above about 500 miles).  In the event that the tether fails, a rotating tether will make it more difficult to prevent the counterbalance station from re-entering the atmosphere, and more difficult for any docked launch vehicles to re-enter safely.  The rotation would complicate the use of a momentum engine, since the engine must not rotate during operation. Also, rendezvous operations will happen at higher acceleration (and therefore will be more difficult) for a rotating tether.

The Gondola:
Since the mass of the track is the main determining factor for the tether mass, the gondola might be sized to carry a only third as many passengers as the launch vehicle; perhaps 35. Since the gondola ride will likely last several hours, the gondola will have a galley, lavatory, etc. In addition to the main door, emergency exits (including pop-out tunnels) will be provided to allow evacuation to another gondola on an adjacent track.

The Motor Train:
To minimize the required mass of the track, the gondola's weight will be spread over a couple of miles of track, by a long "motor train". This reduces the lift requirement of the system to a couple of pounds per linear foot (with 12 lbs/sqft of radiation shielding, the gondola+motor train will likely have a mass of 50,000 lbs and the maximum Gee load is about 1/3). The lateral pressure on the track will be similar (assuming 1000 ft or so of cable between the gondola and the motor train, the coriolis force will dominate: coriolis acceleration is 0.0043 Gees at 100 m.p.h.).

Even with the track optimized for small passenger gondolas, it should still be possible to carry the full cargo load of the launch vehicles on one cargo gondola, by using an extra long motor train, and running at slower speeds (the cargo gondolas would not be pressurized or shielded and therefore would have less structural mass).

Note that the gondolas hang beneath the motor train. As a result, they cannot go past the zero Gee point. The passengers traveling from the dock to the counter balance station must must stop at the zero Gee point and change from a Earthward-hanging gondolas to one hanging away from Earth.

The motor train will have several different types of cars. The regular ones may be a few dozen feet long, and form the bulk of the train. The first and last cars are extra long, and have a rigid frame. The first car must be able to push upward past a section of track that has lost power or has a motor system failure (to improve system reliability). The last car carries extra lateral load, due to the gondola cable.

An independent track inspection car will travel ahead of the main train to detect damaged track and allow the train to stop and avoid derailing.

Ballast cars (which aren't part of the gondola's motor train) could be stationed along the length of the elevator for energy storage. They move in a direction opposite to that of the passing gondolas (a rising car consumes electrical energy and a descending one generates electrical energy). The combined mass of the ballast cars will be large (perhaps 10 to 50x the mass of one gondola), so that only cars within one hundred or so miles of a gondola will transmit power to the gondola. The use of ballast cars reduces the electrical efficiency of the system, but eliminates the need to schedule upward passenger service at the exact same time as the downward service (although the winch that reels in the launch vehicle will generally operate simultaneous to the gondolas), and smoothes out the peaks in the electrical load on the solar electric panels (as well as allowing operation during the nighttime side of the orbit). A ballast car moving along the lowermost one hundred miles of track will store 125 Watt-hours per kg of mass, which is about the same energy in a kg of chemical batteries, so that for this tether length it's better to use batteries.

The tether will support three parallel monorail gondola tracks; two for simultaneous bidirectional travel, and a third track for ballast cars (which also serves as a backup and allows parking of maintenance vehicles). Track switches will be provided at perhaps a dozen or more places along the tether so that gondolas can move from one track to another to route around a damaged section or to simply change direction at the end of a run.

Electric Motor-Generators:
To allow descending gondolas to provide energy for ascending ones, the motor trains must be powered by electric motor-generators and be connected somehow to the tether's electrical wires. To allow very light weight gondola track to be used, extremely numerous small wheels will be used to spread the vertical load over a long span of small track, perhaps one set of drive wheels for each foot of track. To minimize transmission losses, each pair of wheels should have its own drive motor (totaling perhaps ten thousand motors for each train).

The motors in the motor train must have high power to mass ratio. For a modest 50 m.p.h. climb rate, the overall ratio (including payload) is 30 Watts/pound (88 hp/ton), but the motors alone must be several times better, perhaps 300 W/lb (this is normal for the 100 kW motors made for electric automobiles, but challenging for the small 100 W motors used here²⁵ ). To achieve the necessary high power to mass ratio, a high speed (above 8,000 R.P.M.) brushless motor would be used.

In particular, the 3-phase AC induction motor is a good fit. In addition to not having brushes (which wear out), they have the nice property of being able to run at a speed which can vary by a few percent, while using a fixed-frequency power source. This allows many of the motors to share the same AC power bus, and still have good mechanical load sharing. In contrast, brushless permanent magnet motors would require individual variable frequency drive circuits because slight differences in wheel diameter would force each motor to run at a slightly different frequency (electronic drive circuits are necessary in any case, since the input power will likely be DC, but a dozen or more induction motors could share the same drive circuit).

Even though the motors are brushless, the trains must have brushes to electrically connect to the electrical rails on the track. Because the train has power cables running its length, it can easily move past a long span of track that has lost power.

In a more advanced system which avoids the need for a direct electrical connection between the gondola and track (which therefore allows higher speed gondolas), the track could serve as the stationary portion of a linear synchronous AC motor. This is to say that the track has flat electric coils (with their axes oriented horizontally) placed in overlapping positions along its entire length. The gondolas’ motors are simply sets of permanent magnets arranged to set up a field across the coils, and therefore provide thrust when the coils are energized. The coils are energized sequentially, at a rate proportional to the gondola speed (i.e. a variable frequency drive circuit is required for the track).

When coupled with a “mag-lev” system to hold the gondolas on the track, high speed operation could be possible. However, for simplicity, wheels could be used, which might allow speeds around 100 to 200 m.p.h.. This simplifies the emergency/parking brake also (the mag-lev system would have required rocket brakes).

Electric Transmission System:
High voltage electrical cables running the length of the tether transmit DC power for two applications: high power between the gondola and the nearest storage battery (about 2.2 MWatts), and a lower level to the batteries from the solar panels. Some of the solar panels will probably be located at the zero Gee point, where the sun shines 69% of the time⁹ and sun-tracking is easy. But most of the panels would be spread along the length of the tether (because the saving in mass for having all the solar panels at zero gee would not offset the mass of extra cables required to send the extra power down the tether). The cables must be insulated since otherwise the sparse plasma that is present in low Earth orbit would leak excessive amounts of current. Electronic power modules placed every 50 feet or so along the tether tap into the high voltage cables and provide conditioned low voltage (200-500V) direct current power to the track.

For a given power requirement, the transmission voltage selection is a compromise between high voltage (which implies lower current and therefore thinner wires) and low voltage (which allows less insulation). At 60 kV, the total current is 36 Amps.

Terrestrial power systems almost always use alternating current, largely due to the favorable cost of AC voltage converters (i.e. transformers). This system uses DC, in part because of the substantial weight savings of DC voltage converters (they actually contain AC transformers, as well as electronic DC to AC inverters and rectifiers, but the operating frequency is very high, which reduces the size and weight). It should also be noted that for long distance power transmission with DC (or short distances with AC), the ratio of voltage to current (the "load resistance") can be arbitrarily selected. This is a big advantage over long distance AC systems, in which the transmission cable's "characteristic impedance" dictates the ratio (the characteristic impedance depends on cable construction, but is never higher than a few hundred to one; a terrestrial 2.2 MegaWatt AC system must use about 55 Amps with air insulated wire pairs or 210 Amps in coaxial cable; power loss in the cable increases with the square of current). The wire's characteristic impedance is normally not an issue for distances under a few miles, but in this case the motors use AC current of higher than normal frequency (probably between 260 and 2000 Hz) because of their high R.P.M.; this makes long distance effects set in sooner.

Another factor that favors DC current over AC current is the problem of insulation. Electrical cables that are negative with respect to the surrounding plasma seem to be easier to insulate than positive (or alternating polarity) cables²⁶ . So here all cables are assumed to be biased negatively. With a DC system then, only half the cables will carry a high voltage, the others acts as ground returns and can have thin insulation. The DC system will use solid core wire; high current AC systems must use hollow wire (which increases the insulation area and therefore adds mass) due to the skin effect (alternating currents tend to flow only at the surface of conductors, so increased thickness has a diminishing bennefit). An AC cable also requires 41% thicker insulation than a DC cable of the same average voltage, due to the peak-to-average ratio of the sinusoidal waveform.

In order to reduce the high transmission voltage to a level more convenient for use, AC systems use transformers. Instead, the DC system suggested here uses a simple series connection of the track power modules to form constant current loops. In this way, each module is designed to handle from 0V (no load) to about 600V (full load), and the total voltage across 100 active modules is 60kV. Inactive modules carry the full current, but have 0V drop across them. The modules must provide 60kV input-to-output isolation, but most of the electrical components don't have the full 60kV of stress on them (which eliminates a lot of high voltage insulation and allows conventional transistor circuits to be used). Each current loop is powered by a battery with a constant current regulator, and powers a dozen or so miles of track.

The electrical transmission system is then composed of a number of constant current loops, place end to end (with circuitry provided to convey power from one loop to the next). The high current level is only transmitted on the loops that power track with a gondola on it. About 30-40% as much current can be transmitted the whole length of the tether to recharge the batteries (with lower efficiency due to the greater distance).

The power cables will be moving quire rapidly through the Earth's magnetic field. This induces a "common mode" (i.e. adding equally to all wires in the circuit) DC voltage in the cables which amounts to a maximum of 314 V/mile at the dock 238 V/mile at the zero Gee area (for 168 kV total over an altitude span of 160 to 775 miles). To keep this from dominating the insulation budget, this Geomag DC will be removed with DC-to-DC converters, spaced along the cables so that the accumulated Geomag DC is never more than 5kV (minimum converter spacing is then 16 miles). The DC current from the batteries is exactly balanced between the High Voltage and the return lines, so no net magnetic thrust or drag is produced. However, the Geomag voltage will tend to produce drag if any current leaks into the surrounding plasma. It might also be desirable to pump DC current into the space plasma (which imbalances the current in the power cables) in order to push against the Earth’s magnetic field and produce thrust.

Transmission System Parameters

Motor train output power: 1.58 MegaWatts (2,100 Hp)	for train mass=50,000 lbs, and 50 m.p.h. climb rate
Input power: 2.17 MegaWatts	for 3% wire loss, 25% other loss
Current Loop Length (low altitude): 16 mile	for batteries placed every 16 miles along tether
System mass: - conductors: 647 lbs/mile - insulation: 1,560 lbs/mile - Pwr modules: 1,860 lbs/mile (@ 500 Watts/lbs) Total: 4,070 lbs/mile= 0.77 lbs/foot	for 60,000VDC + <5,000 VDC-Geomag, 4 HV wires: 3.28 mm dia. aluminum wire w/ 2.51 mm thick Teflon insulation, and 4 return wires 3.28 mm dia, with 0.36mm insulation (insul. padded 5kV+20% on 33kV/mm, wire padded 15%). Each conductor carries 9.1 Amps, & has 99 Ohms
Max power at 600 miles: 660 kiloWatts	maximum power that can be transmitted to dock from zero-Gee area; at 50% power loss (in wires and in anti-Geomag voltage removal).

The wire budget allows four high voltage wires and four low voltage ground returns. For reliability, this might be arranged as two independent current loops, with each loop made of a double wire. The double wire would have cross connects to allow continuity in the event of cable cuts by micrometeoroids.

Momentum Engine:
The momentum engine is required somewhere along the length of the tether (a few dozen miles below the zero gravity point would allow light-weight construction, and still allow liquids to settle) in order to offset drag and the momentum transferred to payloads as they move up the elevator, to the extent that is not offset by downward moving payloads (i.e. during construction of the hotel, or for cargo released into orbit). In this way, the elevator with engine acts as an upper stage rocket, except that more fuel efficient engines can be used.

The thrust required is relatively low. The upper atmosphere will exert about 13 lbs of drag on the the main structure (this varies a lot with design and upper atmospheric temperature)¹⁰. When the docking grapple is extended down to 90 miles, it will contribute another 30 lbs (perhaps 10 lbs average with 4 dockings per day). If cargo is carried to the zero Gee point at the rate of 100,000 lbs per day, another 84 lbs of average thrust is required¹¹. The thrust must be applied symmetrically about the orbit, so if thrust is not available during the 31.6% of the orbit that is not in direct sunlight (775 miles up¹²), then the engine must be shut down another 31.6% of the time on the sunlight side of the orbit to balance it out. The engine is then only usable for 37% of the orbit and must therefore have a thrust of 289 lbs when on to average 107 lbs overall.

For most applications rockets engines are chemically fueled (i.e. the propellant is also the energy source), such as the hydrogen-oxygen engines used in the Space Shuttle. Chemical engines have high thrust for a given engine weight, but have poor fuel economy, or “specific impulse” (specific impulse is the length of time that 1 lb of propellant can produce 1 lb of thrust, which is directly proportional to the exhaust velocity; the Shuttle’s main engines have a specific impulse or Isp of 455 seconds). To reduce the propellant required for a given amount of momentum, several types of engines are feasible which would use energy from the sun to the drive the propellant to higher exhaust velocities. These solar engines can’t be used for launch vehicles since they tend to be many orders of magnitude away from being able to lift their own weight. For the momentum engine on a space elevator, though, low thrust to weight ratio is tolerable, so a solar engine would be the most cost effective.

The solar engines with the best specific impulses utilize solar voltaic panels to produce electricity which is then used in various electrostatic or electromagnetic propellant acceleration devices. They can have tens to hundreds of times better specific impulse than chemical engines, but it takes a lot of electricity to produce a little thrust (for Isp=3,000 seconds and 66% efficiency, each pound of thrust produced takes about 100 kWatts or about 5,000 lbs of solar panels; the power requirement goes up proportionally with Isp).

For one-way cargo transport, the energy cost of associated with an electric momentum engine would dominate the electricity budget. For each (metric) ton of cargo brought to the zero-Gee area, the energy requirement is 400 kWatt*hours to lift the cargo up the tether, and 4400 kWatt*hours for momentum with Isp=3,000 seconds.

An intermediate alternative is solar-thermal engines which have about half the propellant consumption of chemical engines (Isp = 700-1000 seconds), but would use much smaller (i.e. lighter and less expensive) solar collectors for a given thrust than electric engines. Solar thermal engines don’t involve electricity (except for the fuel pump which could potentially be electric), but simply use concentrated sunlight to heat high pressure liquid hydrogen which is then exhausted from a normal rocket nozzle. Hydrogen is the propellant of choice because for a given temperature, hydrogen molecules move the fastest, and therefore can produce the fastest moving exhaust stream.

It might make sense to use enough ion thrusters to offset the atmospheric drag and allow limited cargo capability, and supplement with solar thermals during hotel construction when the cargo rate would require a much higher thrust level. When the solar thermal engines are no longer required for thrust, their solar collectors can be used to provide additional electricity by focusing their light on a set of thermionic diodes or a gas turbine generator.

NASA has demonstrated the use of a solar electric (electrostatic ion accelerator) engine on its Deep Space 1 spacecraft (launched in October 1998). Ion engines have also been used for station keeping on several communications satellites built by Hughes Space and Communications. A solar thermal engine has been lab tested, but not flown¹⁹.

There is a new type of thruster which may be useable in this application called an electrodynamic tether, which could have even lower propellent consumption than an ion thruster, and lower power consumption too. It would utilize a DC current along the tether to push against Earth's magnetic field. The current path through the tether is completed by contacting the sparse but adequately conductive plasma that is present at lower altitudes. Contact can be made with a large area of bare metal on one end, and an electron gun or low powered ion thruster on the other end. For the case of the space elevator, the main electrical power bus would also carry the DC thrust current (the net current flow is just an inbalance in the positive and negative cable currents). A flight experiment to demonstrate the use of electrodynamic propulsion is in the planning stages.²⁴.

Launch Sites and Schedules:
The hotel and space elevator will be placed in an orbit whose period is a sub-multiple of 24 hours (i.e. an integer number of orbits per day - 13 in this case), so that it will pass over the launch sites at the same time from day to day (the exact time will advance through the day over the course of several month, due to orbital precession however). If the launcher has just a small amount of cross-range capability and the launch site is at a latitude nearly equal to the orbit inclination, then the orbit will nearly pass over the launch site in two consecutive orbits each day, so that two launches and two landings are possible from each launch site, daily. Of course, the launch sites will be better utilized if they also serve other hotels, which are also placed in 1.845 hour orbits, but pass over the launch sites at different times of the day. And the space elevator will be better utilized if it is served by several launch sites, and therefore carries more people.

It should be noted that with RLVs (Reusable Launch Vehicles) there is no reason to locate launch sites on a ocean coast. In fact, flying over land provides superior safety since several emergency landing sites can be provided for use during aborts. This is especially important, since the launch vehicles can be expected to occasion fail to dock with space elevator, and therefore a landing site would be required about 2,000 miles downrange from the launch site (this assumes the docking altitude is 90 miles and a low-lift re-entry profile is used; with a high lift X-33 like re-entry profile, a lower dock altitude might be required for the 2,000 mile landing). Depending on the launchers cross range capability, different alternate landing sites might be required for the two daily launches.

During the test flights of the launch vehicle, if flights over populated areas are not desired, then it could be launched from Cape Canaveral in Florida, and land on a barge at sea, but only if it can take off and land vertically (e.g. the Delta Clipper or Roton).

After the launcher is approved for flights over land, Cape Canaveral can continue to be used for daily cargo launches, with the sea barge serving only as an emergency landing site (this might not be cost effective).

With an orbital inclination of 40 degrees or so, the orbit will only pass over Cape Canaveral once per day. The passenger launches can occur at higher latitudes, so that two overflights (and therefore eventually two launch-landing pairs) can occur each day. Passenger launches could occur initially from a launch site in Nevada (with an emergency landing site in Virginia). Later, launch sites could be added in Spain (with an emergency landing site in Turkey) and China.

Future generations of space elevators may gain an economic advantage by being placed above the equator. This would allow launches as often as once per orbit, from a South American launch site. This Mach 19.5 elevator might not allow the launch vehicle sufficient space to abort on continent though.

Construction:

Elevator Construction:
The space elevator will be constructed in low Earth orbit, with the construction equipment located at the high point of the structure. The Earthward end is made first, and will get farther from the construction equipment as new length is added. A mobile loading dock will be frequently reposition along the elevator so it can stay near the zero gravity point (this allows the cargo ships to dock in zero gravity).

During the construction of the first space elevator, if the Reusable Launch Vehicle (RLV) does not have a reasonable payload capacity in Single Stage to Orbit mode, it could be used in a two-stage mode, and release the payloads attached to a booster engine in sub-orbital trajectories. The RLV would then land about 1,000-2,000 miles downrange, and be ferried back to the launch site atop a jumbo jet (as is currently done when the Space Shuttle uses its California landing site).

As an alternative to an upper stage that must be either discarded or returned to Earth for reuse, a small (30x payload mass) rotating tether facility could be used boost the payloads from the resulting sub-orbital trajectory into the construction orbit (see paper by Carrol¹³).

The partially completed elevator can be used to transport the RLV back to the launch site as soon as it’s long enough (and therefore the dock is moving slowing enough) to give the RLV adequate payload capacity. When the space elevator is completed, hotel construction can begin. Also, limited passenger service can begin, with the elevator gondolas serving as makeshift space stations.  

Subsequent space elevators will probably be place in orbits with the same altitude and period as the first one, but different orbital planes. This will allow them to pass over (and be serviced by) the same launch sites, but at different times of day. Construction materials and equipment for these other elevators can be transported to the first elevator. If the materials are attached to a small rocket powered tug, and then released from the tether at a little above the zero-G altitude, they will, over the course of many weeks, drift to the new orbital plane (the orbit planes precess at a rate which is altitude dependent). The rocket tug will then lower them back to the correct zero-G altitude and deliver them to the construction site.

The main tether will probably come off of spools in long, nearly complete segments. A radiation-shielded (but not pressurized) assembly bay would allow space suited workers to complete assembly by adding the rigid trusses, gondola track, and electrical equipment. The assembly bay would be attached to the living quarters for the workers, and would have the solar power arrays and momentum engine that support the construction.

Hotel Construction:
The construction of the hotel could start with either an inflated fabric shell or a rigid shell. The inflated shell would allow a single large structure (as opposed to several smaller rigid modules), which minimized the required mass for radiation shielding. However, the rigid shell will greatly reduce the on orbit assembly labor required by allowing more brackets, cutouts, attachment points, and potentially plumbing to be pre-installed.

Of course the best way to launch a building-sized rigid shell into space is to make the shell double as the propellant tank of a rocket. This idea is as old as Skylab, but is most familiar as a continuously recurring suggestion to use Space Shuttle External Tanks for something, rather than discarding them. In fact, it makes sense to omit the costly Shuttle and its Solid Rocket Boosters, and just attach a reusable engine pack to the External Tank. Surprisingly, current generation ("lightweight") Shuttle tanks only weigh 55,000 lbs empty, so a single Space Shuttle Main Engine (with 470,000 lbs thrust) could carry a partially fueled tank to the space elevator. It probably makes sense to use four engines and launch a full tank directly to orbit, as this would allow extra payload capacity for pre-assembled hotel pieces. The rest of the material required to build the hotel would be sent to the space elevator in cargo versions of the same ships that will carry the passengers.

Construction using the tanks will start with the addition of radiation shielding. This might be in the form of epoxy or other hardenable liquid which is pumped into a specially designed cavity in the tank walls. A shielding thickness of 50 lbs per square foot will add 800,000 lbs to the 154 foot tall tanks (the finished mass of each tank might be two or three times this). After an inflatable airlock is added, construction workers would then have a comfortable "shirt-sleeve" environment in which to work.

Note that the hotel site is in the Van Allen radiation belt, so no "extra-vehicular activity" will be done (at least not in a normal thin-walled space suit).  

Each tank is only large enough for about 50-80 people, so the completed hotel might have a dozen or so such modules split between the zero gravity area and the top of the tether (counterbalance station ). For the counterbalance station, construction would probably occur in the zero gravity area, with special cranes used to move each tank up the tether.

Cost per Passenger:
To obtain an extremely rough estimate of the costs involved, we can use $1 Billion for the cost of the reusable launch vehicle (RLV); this is in between the five of a kind space Shuttle at $1.5 billion, and sixteen of a kind Stealth bomber at $0.8 billion, and well in excess of commercial jumbo jets at under $0.1 billion. The initial deployment might be three space elevator/hotels, one launch site, and fifteen launch vehicles carrying a total of 220,000 passenger per year (several cargo launch vehicles will be required also).

Assuming each RLV launch carries 100 passengers (or 50,000 lbs of cargo, using a 500 lb/per person conversion, to allow for supplies and added structure) and flies 150 times per year, and is amortized over 2 years (the short period compensates for the maintenance cost and return on investment), the per-passenger vehicle cost is $33,333. Alternatively, this corresponds to $67 per pound of cargo (just this portion of the cost).

For the Mach 19.5 dock, a LOX-H2 fueled RLV with dry weight of 12% of gross, would burn 8,800 lbs of fuel per passenger. At the 1990 price of $500/ton, this adds another $2,200 to the per passenger cost and $4.40/lb to the cargo lift cost.

It should also be noted that as the materials are brought from the dock to the hotel site for one way trips, a rocket engine on the space elevator must make up for the momentum gained by the materials; with engines of 800 sec Isp (e.g. solar thermal engines with hydrogen propellant), the propellant for this engine adds 27% to the lift mass (or 20% for materials brought only as high as the zero-Gee point). Including the lift cost for this extra propellant and the elevator cost, the total cost for cargo launches comes to $157/pound.

For the Hotel itself, a crude estimate of the mass can made by taking, say, half the mass of a sea-faring hotel (or cruise ship), which amounts to 30,000 lbs per guest berth¹⁴ (totaling 24 million pounds for 800 guests). Assuming a 4 day stay, and a 3 year amortization period, the hotel mass amortizes to 110 lbs per passenger or $17,700 of lift cost per passenger. We can add another $100/lb as a very rough estimate of material cost (for comparison, sea-faring hotels have a cost in the range of $2/lb and jumbo jets cost around $400/lb), bringing the total amortized hotel cost to $28,600 per passenger.

For the Space Elevator, the Mach 19.5 design made from Thornel T-40 with a safety factory of 2.4 was estimated to have a mass of 34,000,000 lbs. It's capacity is two vehicles per day to and from each launch site (higher for hotels in equatorial orbits). For 100 passengers per vehicle, and one launch sites (a through-put of 200 passenger/day which sizes the hotel to around 800 guests), 365 operating days/year, and a 4 year amortization, the amortized mass is 116 lbs per passenger. Using $157 and $100 for lift and material cost (the current small quantity book price of Thornel T-40 alone is almost this high, but a volume discount will apply and assembly cost should not add much), the amortized Space Elevator cost comes to $30,000 per passenger.

Summing the launch, fuel, hotel, and elevator costs, the total is $94,000 per passenger for a four day stay (and the hotel plus space elevator costs $15 billion). It's likely that the actual price would be subsidized somewhat by revenue from selling high-profit-margin launch services to NASA and satellite companies, and by renting hotel space to movie studios. And of course the costs will come down due to further economies of scale as more elevators, launch vehicles, and hotels are built.


Beyond Low Earth Orbit:
The space elevator can also serve to deploy payloads into Geosynchronous Transfer Orbit (or GTO, an elliptic orbit with a maximum altitude equal to that of the geosynchronous orbit of 22,300 miles), and to deep space (escape velocity) destinations like the Moon, Lagrange points, or Mars. Objects simply dropped from the counter balance station would enter a 1020*3150 mile orbit (i.e. elliptic with 1020 miles minimum and 3150 miles maximum altitude), and could use on-board rockets to obtain the extra 5200 m.p.h. needed to leave orbit (or 3400 m.p.h. for GTO).

It would be much better though to deploy the payloads by first hanging them up to a higher altitude using a deployable hanging tether extended upward from the counter balance station (centrifugal force tends to pull them upward from the counterbalance station). As enumerated in the table following, with the right tether length, the desired departure velocity can be achieve with no further rocket thrust (for round trip travel). The tether mass required for GTO deployments could be accommodated with a large winch.

A potential application of this would be the construction of a pilot plant to demonstrate a geosynchronous orbiting solar power station that beams 100's of megawatts of energy to Earth for conversion to electricity. A cost effective multi-gigawatt production plant would probably have to be built with lunar materials, but a proof of concept plant sent from Earth would be reasonable.

For trans-lunar injection, the tether would be so massive that a winch to carry the full payload capacity of the elevator system might be impractical. However, a fixed tether could be used with a special cable climbing gondola (a Lockheed design using such a system is described at the affordablespaceflight web site²¹). Such a gondola will require a nuclear or solar power source due to the very large climbing energy requirement (1100 Whr/kg from 1020 miles - an order of magnitude more energy than a battery could store), and will be able to achieve very limited climb speeds due to the high power requirement (about 18 KW/ton at 10 m.p.h.). Assuming 50,000 pounds each for the payload and gondola, this cable would mass about 5,000,000 pounds, but would displace 8,800,000 pounds of counterbalance mass (almost half of the total at 1020 miles) due to its greater average distance from the center of gravity. This cable climbing system will probably be too slow for tourism applications (several days one way). And for cargo, it won't generally be cost effective (compared to using a GTO deployment tether with an additional 1940 m.p.h. rocket burn) until a lighter tether material becomes available. While the GTO deployment tether can pay for its mass in propellent savings in about 35 flights, the lunar injection tether would take 7 times longer.

To improve cargo capacity and allow the winch to handle higher orbits, the large hotel counterbalancing the tether could be replaced with a smaller hotel placed part way up the tether, plus a small dock facility placed even further from the zero Gee point. The maximum dock altitude that can be provided in this way is 1,460 miles with a dock of 1 million pounds mass. It would feel 0.25 Gees. A cable to deploy payloads at escape velocity would then be 475 miles long, and have a mass 8.3x the payload. The GTO cable is only 160 miles long and has mass 1x the payload. But because the GTO cable from a 1/10th Gee Hotel works fine for deployments once per day or less, this scheme (which adds to the cost of the main tether) would only be used if the GTO deployment rate was required to be several times per day.

Direct trajectories to the Moon can only be deployed twice per month (that's how often the Moon will pass through the orbital plane of the space elevator). However, when the travel time is not critical, payloads can be deployed into elliptic parking orbits in advance of the departure window. When the window does open, a small rocket burn would be used to send the payload to the Moon (for example, when deployment velocity is 500 m.p.h. less than lunar injection velocity, the payload will enter an elliptic orbit with an apogee is 74,000 miles, and an orbit period of 2 days). In this way, the deployment tether might have a throughput as high as one payload every two days. This could make it feasible to import construction materials (perhaps dirt and gravel for radiation shielding) from the Moon, for use in low Earth orbit.

It should be noted that a space elevator in an orbit with an inclination matching that of the Moon could provide direct deployments to and from the Moon (or Lagrange points), several time per day. However, this is not practical because of orbital precession ("nodal regression"). The Earth's gravity will pull the space elevator's orbital plane out of the plane of the Moon, and the large mass of the elevator makes it impractical to use station keeping fuel to keep the planes aligned.

Free-Flying Booster Tethers:
For occasional deployments at escape velocity, a free-flying rotating tether could boost payloads from the space elevator to escape velocity, using a much less massive tether (see Tethers Unlimited site¹⁷for their lunar tether system). If the initial tether orbit is highly eccentric, and a large orbit change is allowed, then this type of facility need only be about twice as massive as the payload.

A different type of transfer tether could be built for lunar tourism with departures several times per week (no need to wait for bi-monthly departure windows). The facility's required ballast is much higher due to the frequent departures (perhaps 30-100x the payload mass). Also, the facility would likely use a relatively round orbit at nearly geosynchronous distance (so that departures from the space elevator would use a GTO trajectory). The greater distance to the tether reduces the cost of the change of orbit inclination for the Moon compared to the space elevator (which requires a rocket engine burn) but also increases travel time (from 3 days to 4) and limits the departure rate to one every couple of days. A special type of free flying rotating tether can correct the payloads orbital inclination in an energy conserving way. See A Transfer Tether.

For landing on the Moon, Lunar tourists will use a special tether called a Lunavator (see also Tethers unlimited site). This type of tether is placed in a circular orbit around the Moon, and rotates. The tether length and rotation rate are selected so that the tether tip touches the Lunar surface with zero relative velocity (and can therefore pickup or drop off a payload) as it rotates past. Because the Lunavator conserves momentum, the only propellant requirement of the Lunar landing vehicle is for fine course adjustment.

Mars via L1:
Departures to Mars from the Space Elevator would be somewhat different than launches on expendable rockets¹⁸. A major difference is that expendables can be built to any size, so it's prudent to use one large enough to carry the whole mission. However, in an age of low cost flights with reusable rockets, large missions will likely be made by consolidating payloads from several launches of moderately sized launchers, in Earth orbit. Once a Mars base is established, much of the required cargo can be sent one launcher's worth at a time, but crews will probably still travel in large ships, so that they are never separated from their supplies.

Multiple payloads could be consolidated in a round parking orbit (or attached to the space elevator at the zero-Gee point). However, this would not take advantage of the space elevator's upward deployment capability (therefore the mission still would require a large rocket upperstage to reach injection velocity), and has the added problem that the consolidation site is in the radiation belts.

Alternatively, the payloads could be consolidated in an elliptic parking orbit, which allows the space elevator to provide all but about 500 m.p.h. of escape velocity (as reported by Zubrin¹⁸, the trans-Mars injection velocity is another 2,500 m.p.h. above escape velocity for the "normal" 180 day transit time and 1,100 m.p.h. above escape velocity for the 250 day "cargo" trajectory). But this is only possible if the payloads are all launched within a short period of time. Over a period of weeks however, the parking orbit will precess and therefore become inaccessible to the space elevator. (It would be possible however, to use an elliptic parking orbit if there was a A Transfer Tether in use for other purposes.)

Another issue with the Space Elevator is that it will not generally be in a good orbital plane for direct trans-Mars deployments. Very slow vehicles such as those utilizing solar sails or ion propulsion could correct the plane in route with a small penalty in transit time.

A solution for both the consolation problem and the departure plane problem, would be to park at the first Lagrange point (L1) for payload consolidation. This point, which is about 85% of the way to the Moon shares the Moon's 28 day orbital period, so it can be reached bimonthly from any low Earth orbit, and will pass through the plane of the Mars departure orbit bimonthly. Vehicles parked there have near escape energy and can easily be maneuvered (i.e. a couple hundred m.p.h. delta-V) into a low Lunar and then low Earth flyby. The low Earth flyby is necessarily the location of the final rocket burn for trans-Mars injection (the burn delta-V is about 2,700 m.p.h., but due to the slingshot effect, it provides energy equivalent to a 12,000 m.p.h. burn at L1).

Additionally, the Lagrange points are well beyond the Van Allen radiation belt (unlike the space elevator orbit). This may simplify payload consolidation by allowing astronauts to perform space walks.


Hazards and Safety Concerns:
The space elevator will have a redundant multi-element tether construction (perhaps using a Hoytether or Hoytape design), so the structure can survive many micrometeor impacts. Derelict space craft are a big concern, since one could easily sever the entire tether. Therefore traffic control will be necessary over the whole range of altitudes occupied by the space elevator: from 160 miles to 1,600 miles for the initial design, and perhaps to 2,300 miles for second generation elevators (i.e. all spacecraft and satellites in this range must yield right of way to the space elevators). Also, all satellites will have to be removed from orbit at the end of their useful service lives (i.e. end the practice of dumping old spacecraft in junk yard orbits).

With proper design, it can be insured that none of the hotel stations fall to Earth if the tether is cut. For the 860 mile long tether discussed here, the lower most hotel station must be at least 670 miles above the Earth (at this point, 105 miles below the zero Gee point, there is 5% of Earth normal gravity). Escape pods released at or above this altitude will enter a stable orbit at least 200 miles above the surface.

If a tether cut occurs below this altitude, the hotel complex will be safe (although the lower part of the space elevator, including the dock, will be lost). If a tether cut occurs above this altitude, and above a hotel station, then the tether must be deliberately severed just below the lowest hotel station, probably using explosives, to insure that no part of the hotel is pulled down with the severed tether segment.

It should be possible to design the gondolas to include a provision for loss of suspension (either failure of the main tether, failure of the tether between the gondola and motor train, or failure of the motor train). One option would be use of a solid rocket motor to boost a separated gondola to orbital velocity. For the Mach 19.5 tether described here, the solid propellant would have to be 44 percent of the gondola mass.

An option that scales better to longer (and slower) tethers would implement a controlled atmospheric re-entry. The gondola must have a massive outer skin for radiation shielding. This skin could be designed to double as an ablative heat shield for re-entry. A parachute would cushion the landing. Depending on the shape of the gondola, the atmospheric re-entry could result in dangerously rapid deceleration.

Orbital Crowding:
Due to its enormous size, the space elevator is not a good neighbor for the haphazardly placed collection of satellites in low Earth orbit today. The risk of collision is greater though, for other tethers and space elevators. It will probably be simplest if all permanently deployed tethers and space elevators share an identical orbital period, and maintain a fixed spacing with respect to one another. Of course, after some period of time, it will be desirable to move to the next generation of longer, lower docking-velocity space elevators. In this case existing structures would be renovated (i.e. have their main tethers replaced).

When a launch (or Lunar excursion) vehicle docks (or undocks), the orbit of the space elevator will shift somewhat due to the new center of gravity. But in this case, the structure is more than two orders of magnitude more massive than the vehicle, so the effect is tolerable. The perigee of the space elevator will dip by about 5 miles, and it will move away from its station in the orbit by about one degree per hour. The momentum engine can be fired to adjust for situations when launch vehicles don’t dock or depart when expected. This should allow one or two dozen elevators to share an orbit (more with larger and heavier space hotels).

Ultimately, the upper limit on the number of space elevators that can orbit the Earth at the same time might be probability that one severed tether could sever others. When a hanging tether is severed in the middle, the lower half re-enters the atmosphere, and the upper half will have an elliptical orbit with a slower period. The upper severed fragment must then be de-orbited or lowered into an unoccupied slot before it has a chance to collide with (and sever) another space elevator.

Practical Concerns:
The most obvious objections to the whole scenario are that it’s too expensive and/or too dangerous. It’s certainly true that past space programs have been expensive and dangerous, but the same was true of the early aviation industry.

In any high technology design, new and immature products will have design defects. Fixing them takes time and engineering resources, so the thoroughness with which these defects are removed depends on market demands. A government sponsored launch system that only makes a handful of flights a year could never justify the low defect rates that a commercial airliner has (or that a commercial space vehicle would require).

Similarly, any high technology design will be expensive to produce in “onesy-twosy” quantities. Production costs and purchase price inevitable come down as production volume go up. For anything built on a government contract in particular, the supplier never has an incentive to cost optimize. A more prudent strategy is to make an attractive contract proposal, and then raise the price later by exploiting contract loop holes and fine print. Of course in the commercial world, market acceptance of a product or service is always tied to meeting cost targets, so the supplier is much more motivated to consider cost.

Radiation Concerns:
Unfortunately, the Van Allen radiation belt starts as soon as the atmosphere stops (at least in the polar regions), in fact it’s considerably worse at the hotel altitude than it is near the dock. The hotel will need a rather massive radiation shield (with perhaps 50 lbs of shielding for every square foot of outer wall, for example with aluminum 3.6 inch thick, which would amount to 20-50% of the hotel mass). Even so, early hotels will probably have a high enough radiation level that it would be prudent to keep crew-member tours of duties down to a few months.

The elevator ride will be brief enough to allow only moderate shielding, perhaps 12 lbs/sqft , but even this amount will constitute the majority of the mass of the gondola.  This probably means the elevator gondolas will not carry a crew to limit routine crew-member exposure. The design limit will specify that even when mechanical problems strand passengers in the gondola for a day or two, they wouldn’t have a high enough dosage to suffer radiation sickness.

The radiation will be an inconvenience during hotel construction, since the construction crew will not be able use space suits to do space walk construction. But this is not such a big issue since that's an inefficient way to do construction anyway (compared to working in a pressurized assembly area).

The following graph shows radiation dosage as a function of altitude¹⁶. For reference, a medical CAT scan exposes patients to 5-10 rads, 300 rad-equivalent is considered lethal, and 25 rad-equivalent/month is the NASA limit for astronauts. The data shows that a 3g/cm^2 shield corresponding to 6.1 lbs/sqft, is not adequate (note 10^4 rads/year= 27 rads/day), but does provide about a 10x reduction in exposure. Extrapolation suggests that the dosage might be as low as a couple of rads per day with 12 lbs/sqft in the zero-Gee area (and 18 lbs/sqft at the counterbalance station).

The shielding levels assumed here for early hotels is much less than has been discussed for a Lunar base (NASA has talked about 1000-2000 lbs per square foot), in part because the Earth’s magnetic field protects the hotel from most of the radiation coming from our Sun and some of the energy from cosmic rays (except in the polar regions). The heavy Lunar shielding is probably beyond the point of diminishing returns for low Earth orbit, and reflects the great difficulty in getting the dosage rate below about 1 rad per month.

Radiation will also shorten the life of the solar panels, and restrict the selection of tether materials. Spectra-2000 fiber is said to be strongest available today (for its weight), however, like most plastics, it is not radiation tolerant, and therefore will not have acceptable design life at the higher altitudes. Above 400-500 miles, the tether can be constructed of graphite fiber, which will loose strength at the rate of only a few percent per decade. The zero gravity solar panels will probably have to be replaced every 5-10 years (but panels near the dock should last much longer).

Conclusion:

Impact on Government Space Programs:
Most NASA programs would surely evolve to make use of the space elevator facility. In particular, the focus of NASA's manned space program is likely to move beyond low Earth orbit, to the Moon, Mars, or the asteroid belt. Lunar development in particular could reduce the cost of large construction projects in Earth orbit by supplying materials for lower launch costs ¹⁵.

The International Space Station might get relocated to Lagrange point One (L1), where it could serve as a staging area for missions to a Mars base. NASA astronauts would then book passage on a commercial flight to the space hotel, then catch a space taxi over to their space station.


How to Make It Happen:
The US government must nurture this new industry, in the same way it supported the communications satellite industry in the 1970’s, the aircraft industry in the 1920’s and the railroads before that.

NASA should re-align it’s goals for the manned space program to specifically support this effort. In particular, tether demonstration flight should be raised to a high priority. The X-33 program can be used to demonstrate and test engines and thermal protection systems, and maybe even provide a design that could be developed into a commercial launch vehicle.

In addition to the X-33 (VentureStar) program, NASA should also resume research into the vertical take off and landing Delta Clipper launch vehicle. This configuration has a potential safety advantage of allowing landings with full or nearly full fuel tanks for an early launch abort capability that a glider-landing vehicle like the X-33 does not have. And if a low-fuel consumption idle mode is provide, a vertical landing vehicle could leave its engines running from launch to docking (and from un-docking to landing), to eliminate concerns about a critical air-start during atmospheric re-entry for landings and aborts (it only takes about five minutes to get from the dock to the landing site).

Lastly, we should adopt firm policies against leaving man made debris in orbit, ranging from old satellites to paint flecks. This particularly important given the recent trend toward deployment of communication satellites in low Earth orbit, in vast constellations involving dozen of satellites per system (e.g. Globalstar, Iridium, & Teledesic), and renewed interest in space based weapons.

One way to give teeth to such an end of life de-orbit policy would be to require that a cash de-orbit deposit (amounting to say 2-5% of the launch cost) be held in escrow until the satellite is removed from orbit. If the satellite's de-orbit system fails, the de-orbit deposit (along with accumulated interest) would then act as a bounty to compensate space trash removal companies. We could also require all satellites to have a standardized docking mechanism, so if they are inadvertently abandoned in orbit, an automated craft can later dock with and de-orbit it.

About the Author:
Nathan Wilson has dabbled in amateur “rocket science” since being inspired several years ago by the San Diego L5 society. He holds a Master of Science degree in electrical engineering from Stanford University, and works as an electrical engineer for Seagate Technology.

Comments and/or questions about this document may be addressed to Nathan2go@aol.com.

References, Notes, & Websites:
1, 21. Tether background/history:
Space Towers  gives a Space Elevator history, including the work of Tsiolkovsky, Y.N. Artsutanov, and Jerome Pearson.
http://liftoff.msfc.nasa.gov/academy/tether/spacetowers.html
Advanced Space Transportation Program: Tethers - Marshal Spaceflight Center, includes their tether projects.
http://std.msfc.nasa.gov/astp/tethers_main.html

affordablespaceflight
http://www.affordablespaceflight.com
This site, dedicated to making spaceflight affordable to the individual, reports on a 1994 study Lockheed did for NASA, with a space elevator and a scramjet powered horizontal takeoff and landing launch vehicle. The Lockheed elevator design is much less massive than the one described here, but it lacks the capability to carry the launch vehicle back to the launch site and requires a nuclear powered cable climber. return to text 21

The Rocket/Skyhook Combination
http://www.frc.ri.cmu.edu/~hpm/project.archive/1976.skyhook/1982.articles/carley.mss
by F. Burke Carley and Hans P. Moravec, this was published as part of a 1983 workshop proceedings.  This is also distributed as part of a U.S. government micro fiche collection, NAS 1.26:171197.  It gives an overview of  tether types, including vertical, rotating, and rocking, and suggests an implementation with a Kevlar tether and the Space Shuttle as a launcher.

Space Elevators
http://www.frc.ri.cmu.edu/~hpm/project.archive/1976.skyhook/1982.articles/elevate.800322
This 1980 article by H. Moravec and R. Forward describes the concepts behind rotating and non-rotating Earth to orbit space elevators.
return to text1

2. SSTO calculations by the Author (SSTO spreadsheet) have been verified against published data from McDonnell Douglas' Delta Clipper program, which proposed a 1.28 million pounds gross vehicle, with 90% propellent fraction (including the requirements for a verticle landing), and 1.56% payload fraction.

Once a vehicle has transitioned from verticle to horizontal flight, the amount of propellant required for a given velocity change is given by the so called Rocket equation: Mr = 2.718^(velocity_change/Isp/g), where Mr is the ratio of the mass before the propellant is burned to the mass afterward. The propellant fraction is therefore = (Mr-1)/Mr. g is a conversion factor equal to 21.8 m.p.h. per second (=9.8 m/s/s). Isp is a propellant/engine characteristic called specific impulse, which is about 450 seconds for hydrogen-oxygen and 300 seconds for kerosene-oxygen.
return to text 2

3, 7, 17. Other tether data from Tethers Unlimited, inc.
Tethers Unlimited Home Page
http://www.tethers.com
7. The Hoytether:  A Failsafe Multiline Tether for Long Tether Lifetimes
http://www.tethers.com/Hoytether.html
17. Tether Transportation for LEO to the Lunar Surface
http://www.tethers.com/LEO-Lunar.html
return to text 3  return to text 7  return to text 17

4,8. Carbon Fiber radiation tolerance data came from, “The Effects of Space Radiation on a Chemically Modified Graphite-Epoxy Composite Material” by Reed, Herakovich, and Sykes. NASA Grant NAG-1-343, October 1986. Microfiche NAS 1.5:89232. They irradiated test samples with a 1.0 MeV electron beam for 1*10^10 rads and noted no degradation in the graphite fiber, but the epoxy turned to mush. It should be noted that a small fraction of cosmic rays, unlike the test beam, does have sufficient energy to cause atom smashing to the carbon fibers.
return to text4  return to text8

5. Carbon Fiber data is from Amoco Performance Products, Inc. 4500 McGinnis Ferry Road, Alpharetta, GA 30202-9893, 800-222-2448.
Amoco Chemical/Thornel
http://www.amocochem.com/altsite/products/thornel/thornel.htm
return to text5

6, 11. Calculations by the Author (Sp Elev spreadsheet).
SPELEV5B.XLS in Excel 5.0 format.
SPELEV_SS.CLW in Clarisworks 1.0.
SPELEV_SS.CLW4 in Clarisworks 4.0.
return to text6  return to text11

9, 12. The fraction of an orbit that a satellite is in the Earth’s shadow is given by:
(arcsin((radius_Earth)/(radius_Earth+altitude_satellite))/(180 degrees))
which is 31.6% at an altitude of 780 miles.
return to text9  return to text12

10. Aerodynamic drag for the tether was calculated with the equation and atmospheric data from “The Tethers in Space Handbook”. The tether was assumed to have 72 separate cables, and the three gondola tracks were assumed to have a combined exposed width of 200 mm.

“The Tethers in Space Handbook” is available for download from: (this also includes links to other tether related sites)
AIRSEDS Info
http://www.airseds.com/tethersinfo.html
return to text10

13. “Preliminary Design for a 1 Km/sec Tether Transport Facility”, by Joseph A. Carroll of Tether Applications. A paper presented at the NASA OAST Third Annual Advanced Propulsion Workshop at JPL, January 1992. Carroll describes an orbital facility with a 180 mile long tether which captures payloads from the Space Shuttle in a sub-orbital flight. The tether is reeled-in to avoid the need for an elevator.
A summary is given here:
TU Momentum Exchange Tethers
http://www.tethers.com/M-ETethers.html
return to text13

14. Data on Seafaring Hotels is from: Forbes magazine, 7 July 1997, “Floating Monstrosities” which describes Carnival’s new cruise ship Destiny, which was at that time, the world’s largest at 101,000 tons. It carries 3400 passengers and costs $400 million.
return to text14

15. For a general introduction to why it’s desirable, prudent, and feasible for humans to go into space, see Gerard K. O’Neill’s book about human colonies in space, “The High Frontier”. 1976 original edition. It also discusses collection of solar energy in space for use on Earth and mining of lunar resources for construction in Earth orbit.
return to text15

16. Radiation data is from JPL Publication 83-26, “Radiation Protection for Manned Space Activities” by Thomas Jordan. This is distributed as part of a U.S. government micro fiche collection, NAS 1.26:173202. The radiation graph is sub-referenced to H.R. Rugge, “Space Radiation and Its Possible Effects on Man in Space”, from Space Systems and Technology Workshop II, Sept. 1982.
return to text16

18. Robert Zubrin provides an excellent discussion of Mars exploration issues in his 1996 book, “The Case For Mars, The Plan to Settle the Red Planet and Why We Must”. 1976 original edition. There is also a companion web site: Marsdirect Homepage http://www.nw.net/mars/marsdirect.html.
return to text18

19. Some information on an experimental solar thermal rocket engine is available at: SOTV.html http://www.boeing.com/defense-space/space/propul/SOTV.html.

NASA's Deep Space One spacecraft and its ion engine are describe at:
ds1-index.html http://nmp.jpl.nasa.gov/ds1/index.html
return to text19

20, 22, 23. For further reading about single stage reuseable rockets see:

Rotary Rocket Company  http://www.rotaryrocket.com/ - Rotary Rocket Company's Roton single stage to orbit, vertical takeoff, vertical landing, kerosene and oxygen fueled vehicle with fold-away helicopter rotor landing system.

Pioneer Rocketplane  http://www.rocketplane.com/ - Pioneer Rockplane's Pathfinder Mach 15, horizontal takeoff and landing, kerosene and oxygen fueled vehicle with in-flight liquid oxygen loading.

Venturestar homepage   http://www.venturestar.com/ - The NASA's X-33 program and Lockheed-Martin's proposed vertical takeoff, horizontal landing, hydrogen-oxygen fueled SSTO.

the X-33 Notes and comments by the author.
return to text20 return to text22 return to text23

24. The NASA electrodynamic tether propulsion experiment is called:  ProSEDS
http://std.msfc.nasa.gov/astp/proseds.html
Electrodynamic Propulsion
http://std.msfc.nasa.gov/astp/tethers_electetherprop.html
return to text24

25. As examples of existing motor technology, MicroMo corporation's model number 2444 is a 3.5 oz. brushless permanent magnet motor that puts out 37 Watts at 30,000 rpm, which is 164 W/lb. Coreless Motors by MicroMo Electronics http://www.micromo.com/.
return to text25

26. Electrical tether construction is discussed in "Design and Fabrication of the 20 km/ 10 kV Electromechanical Tether for TSS-1 Using High Impact Conductor (Highwire) (TM)" by E. Scala, Lee Marshall, Douglas Bentley. This paper (published in Advances in the Astronautical Sciences Volume 62: Tethers in Space), describes the tether used on TSS-1, the tethered Space Shuttle experiment. This tether used 0.3mm thick FEP (Dupont Teflon) for 10kV insulation strength.
return to text26

You are visitor number:  

Back to Top
Back to Site Index Page