26 August 2016

Catching Some Comet Dust: Giotto II (1985)

Giotto 1 liftoff. Image credit: European Space Agency
On the overcast morning of 2 July 1985, the eleventh Ariane 1 rocket to fly lifted off from the Centre Spatial Guyanais in Kourou, French Guiana, an outpost of the European Community located a few degrees north of the equator on the northeast coast of South America. The last Ariane 1 to fly, it bore aloft Giotto, the first European Space Agency (ESA) interplanetary spacecraft. Giotto's destination was Comet Halley.

A "dirty snowball" containing materials left over from the birth of the Solar System 4.6 billion years ago, Halley needs about 76 years to revolve around the Sun once. Its elliptical orbit takes it from the cold emptiness beyond Neptune to the space between the orbits of Venus and Mercury. Halley travels around the Sun in a retrograde orbit, meaning that it orbits "backwards" relative to the eight planets and most other objects making up the Solar System.

Comet Halley has passed through the inner Solar System 30 times since its first verified recorded apparition in 240 B.C. In 837 A.D., it passed just 5.1 million kilometers from Earth; during that apparition, its dust tail must have spanned nearly half the sky, and its bright coma – the roughly spherical dust and gas cloud surrounding its icy nucleus – may have appeared as large as the full moon.

Shortly after its bright apparition in the year 1301, Italian artist Giotto di Bondone was inspired to add Comet Halley to his painting The Adoration of the Magi. The Giotto spacecraft was named for him.

Comet Halley appears near the top of Giotto di Bondone's The Adoration of the Magi. 
Throughout most of its known apparitions, Comet Halley was not understood to be one comet repeatedly passing through the inner Solar System. Not until 1705 did English polymath Edmond Halley determine that comets seen in 1531, 1607, and 1682 were probably one comet orbiting the Sun. He predicted that, if his hypothesis were correct, the comet should reappear in 1758 (which it subsequently did).

The Ariane 1's third stage injected 980-kilogram Giotto into a 198.5-by-36,000-kilometer orbit about the Earth. Thirty-two hours after launch, as it completed its third orbit, flight controllers in Darmstadt in the Federal Republic of Germany commanded drum-shaped Giotto to ignite its French-built Mage solid-propellant rocket motor. The motor burned 374 kilograms of propellant in 55 seconds to inject the spinning 2.85-meter-tall, 1.85-meter-diameter spacecraft into orbit about the Sun.
 
Two months before Giotto's launch, Americans P. Tsou (Jet Propulsion Laboratory), D. Brownlee (University of Washington), and A. Albee (California Institute of Tech) proposed in a paper in the Journal of the British Interplanetary Society that a second Giotto mission be launched to fly close by one of 13 candidate comets between 1988 and 1994. They proposed that the new spacecraft, which they dubbed Giotto II, might launch on an Ariane 3 or in the payload bay of a Space Shuttle. Giotto II's "free-return" trajectory would take it as close as 80 kilometers from the target comet's nucleus, then would return it to Earth.

Near the comet, Giotto II would expose sample collectors to the dusty cometary environment. Near Earth, it would eject a sample-return capsule based on the proven General Electric (GE) Satellite Recovery Vehicle (SRV) design. The capsule would enter Earth's atmosphere to deliver its precious cargo of comet dust to eager scientists.
 
Tsou, Brownlee, and Albee pointed out that the Mage solid-propellant motor had not been required to boost Giotto into interplanetary space; that is, that the Ariane 1 could have done the job itself. Giotto was, however, based on a British Aerospace-built Geos magnetospheric satellite design, which included the Mage motor. Re-testing the design without the motor would have cost time and money, so ESA elected to retain it for Giotto. After noting that the GE SRV could fit comfortably in the space reserved for the Mage, they proposed that, in Giotto II, the reentry capsule should replace the motor.

Giotto included on its aft end a "Whipple bumper" - named for its inventor, planetary astronomer Fred Whipple - to protect it from hypervelocity dust impacts. During approach to Comet Halley, the spacecraft would turn the bumper toward its direction of flight. The bumper comprised a one-millimeter-thick aluminum shield plate designed to break up, vaporize, and slow impactors, a 25-centimeter empty space, and a 12-millimeter-thick Kevlar sheet to halt the partially vaporized, partially fragmented impactors that penetrated the aluminum shield.
 
In the case of Comet Halley, dust would impact the bumper at up to 68 kilometers per second. Tsou, Brownlee, and Albee noted that the 13 candidate Giotto II target comets were all less dusty and would have lower dust impact velocities than Halley. Because of this, Giotto II would need less shielding than Giotto.

Comet dust would, nevertheless, create challenges for Giotto II. Tsou, Brownlee, and Albee devoted much of their paper to a description of how the spacecraft might successfully capture dust for return to Earth. One proposed capture system, based on the Whipple bumper design, would use a shield made from ultrapure material to vaporize and slow impacting dust particles. The vapor from the impactor and the impacted part of the bumper would then be captured as it condensed. Scientists would disregard the bumper material when they analyzed the condensate.
 
Tsou, Brownlee, and Albee also noted that thermal blankets from the Solar Maximum Mission (SMM) satellite, launched into Earth orbit on 14 February 1980, had demonstrated that intact capture of high-velocity particles was possible. The multilayer Kapton/Mylar blankets, which were returned to Earth on board the Space Shuttle Orbiter Challenger at the end of mission STS 41-C (6-13 April 1984), were found to have collected hundreds of intact meteoroids and human-made orbital debris particles.

The three scientists described preliminary experiments in which gas guns were used to fire meteoroid and glass fragments at "underdense materials," such as polymer foams and fiber felts. The experiments suggested that such materials could capture at least partially intact comet dust particles.

Giotto's encounter with Comet Halley spanned 13-14 March 1986. At closest approach the spacecraft passed just 596 kilometers from Halley's nucleus. The comet's 15-by-eight-by-eight-kilometer heart turned out to be extremely dark, with powerful jets of dust and gas blasting outward into space.

Artist's concept: Giotto at Halley. Image credit: European Space Agency
Halley's hot heart as imaged by ESA's Giotto spacecraft. 
The intrepid probe suffered damage from dust impacts – one large particle sheered off more than half a kilogram of its structure – but most of its instruments continued to operate after the Comet Halley flyby. ESA thus decided to steer Giotto toward another comet.

On 2 July 1990, five years to the day after its launch, Giotto flew past Earth at a distance of 16,300 kilometers, becoming the first interplanetary spacecraft to receive a gravity-assist boost from its homeworld. The gravity-assist flyby put it on course for Comet Grigg-Skjellurup, which it passed at a distance of 200 kilometers on 10 July 1992.
 
After determining that Giotto had less than seven kilograms of hydrazine propellant left on board, ESA turned it off on 23 July 1992. The inert spacecraft flew past Earth a second time at a distance of 219,000 kilometers on 1 July 1999.
 
By that time, a comet coma sample return mission was under way with two of the Giotto II proposers playing central roles. In late 1995, Stardust became the fourth mission selected for NASA's Discovery Program of low-cost robotic missions. Brownlee and Tsou, respectively Stardust Principal Investigator and Deputy Principal Investigator, designed the mission's sample capture system.

Artist's concept of the NASA Stardust spacecraft at Wild 2. Image credit: NASA
The 380-kilogram Stardust spacecraft left Earth on a free-return trajectory on 7 February 1999, and flew past Comet Wild 2 (one of the 13 Giotto II candidates) at a distance of about 200 kilometers on 2 January 2004. Stardust captured dust particles in aerogel, a silica-based material of extremely low density that was invented in the 1930s. Tsou, Brownlee, and Albee had apparently been unaware of aerogel when they proposed Giotto II in 1985.

Stardust returned to Earth on 15 January 2006. Its sample capsule streaked through the pre-dawn sky over the U.S. West Coast before parachuting to a landing on a salt pan in Utah.

When opened on 17 January 2006 at NASA's Johnson Space Center, in the same lab that received the Apollo moon rocks, Stardust's 132 aerogel capture cells were found to contain thousands of intact dust grains captured from Wild 2. Subsequent analysis indicated that some probably formed close to other stars before our Solar System was born.
 
Sources

"Comet Coma Sample Return via Giotto II," P. Tsou, D. Brownlee, and A. Albee, Journal of the British Interplanetary Society, Volume 38, May 1985, pp. 232-239

ESA Remembers the Night of the Comet, European Space Agency, 11 March 2011 (accessed 26 August 2016)

Stardust: NASA's Comet Sample Return Mission, NASA Jet Propulsion Laboratory (accessed 26 August 2016)

More Information

A 1974 Plan for a Slow Flyby of Comet Encke

Cometary Explorer (1973)

Missions to Comet d'Arrest and Asteroid Eros in the 1970s (1966)

12 August 2016

Could the Space Voyages in the Film and Novel "2001: A Space Odyssey" Really Happen? (Part 3)

The glory that is Saturn. The Cassini spacecraft was crossing the plane of the rings as it captured this image, so they are visible only as a sharp thin line running across the center of the image and a complex play of shadows on the planet's northern hemisphere. Image credit: NASA
One can hardly blame Arthur C. Clarke for stubbornly insisting that the spaceship Discovery travel to Saturn. Even with the minimal knowledge of the Saturn system we possessed in 1968, when the film 2001: A Space Odyssey had its debut and the novel by the same name first saw print, it was clear that Saturn is home to some intriguing space oddities which Clarke could put to work in his narrative.

There are, of course, the rings. They make Saturn an austere work of art wrought in ice and orbital mechanics. Clarke attributed their creation to the same advanced aliens who uplifted humankind at the beginning of the book and film 2001, three or four million years before Discovery's launch.

There is also peculiar Iapetus - which Clarke called Japetus (the German spelling) - a 1500-kilometer-diameter moon very dark on one side (its leading hemisphere as it orbits Saturn) and very bright on the other. The arrangement of the dark leading and bright trailing hemispheres mean that Iapetus, Saturn's second-largest moon after Titan, is very bright to Earth-based observers when its 79-day orbit puts it on one side of Saturn and very faint roughly 40 days later, when it moves into view on the other side.

Stanley Kubrick, who directed the film 2001, co-authored its screenplay, and received co-author credit on early editions of the novel, also sought to send Discovery to Saturn, but had to settle for Jupiter. The film's overtaxed art department rebelled - Kubrick, ever the perfectionist, was given to demanding quick-turnaround changes which he subsequently threw away. Perhaps more important, a portrayal of Saturn convincing to 1968 film audiences proved too great a challenge for 2001's pioneering special-effects technology and skilled artisans. Had they known how improbable-seeming Saturn really appears, Kubrick and the production crew might have given themselves a bit of slack.

A raging storm in Saturn's northern hemisphere imaged by the Saturn-orbiting Cassini automated explorer 10 years after 2001. In 1968, observers using Earth-based telescopes believed Saturn's atmosphere to be practically featureless. Image credit: NASA
This is the third and last of a series of posts I have written this summer on real-world proposals for spacecraft and supporting infrastructure meant to emulate the spacecraft and infrastructure portrayed in the film and novel 2001. In the first installment, I described 2001's Earth-moon transport system and a 1997 NASA Lewis Research Center (now called NASA Glenn Research Center) plan to partially duplicate it using the International Space Station, nuclear-thermal rockets, and oxygen mined from the moon.

In the second installment, I described two versions of Discovery, the "hero ship" of 2001. I emphasized the film Discovery; that is, the open-cycle gas-core nuclear-fission design the film's technical consultant, Frederick Ordway, described in the British Interplanetary Society magazine Spaceflight in 1970.

This post picks up the 2001 story where my first post left off, then I will conclude the series by discussing a paper NASA Glenn researchers Craig Williams, Leonard Dudzinski, Stanley Borowski, and Albert Juhasz first presented in July 2001 and subsequently published as a NASA Technical Memorandum (TM) in March 2005. They describe a fusion-powered spacecraft meant to emulate 2001's Discovery spacecraft. They named it Discovery II.

My first post ended with United States Astronautics Agency (USAA) bureaucrat-astronauts accidentally triggering an ancient alarm system. Radio waves blast from an alien black monolith in the moon's great ray crater Tycho. The film 2001 then skips ahead 18 months, and we get our first look at Discovery, her crew, and their daily routine.

As always, Clarke fills in missing details. The novel describes a host of robotic monitors scattered across the Solar System. Each in turn detects the radio signal from the Tycho monolith. Later we learn that data from the monitors enabled scientists on Earth to determine that the signal was directed at Saturn. In the movie, the signal was, of course, beamed at Jupiter.

Fateful decisions are made at the highest level of the United States government. The ancient Tycho monolith and the signal it aimed at Saturn are to be kept secret, allegedly to prevent cultural shock and mass hysteria.

Preparations for Project Jupiter, the first piloted round-trip journey to the Solar System's largest planet, are by this time well advanced; this provides an opportunity. In search of those who received the Tycho monolith signal, the Jupiter ship Discovery will instead travel one-way to Saturn with a gravity-assist flyby at Jupiter to gain speed. I describe the development of gravity-assist spaceflight in my "The Challenge of the Planets" post series; please see the "More Information" links at the bottom of this post.

Discovery's crew of six is split up. A separately trained three-man "survey crew" will travel to Saturn in hibernation; they will thus remain safely incommunicado, ensuring that the mission's true purpose does not slip out during radio communications with Earth. Mission Commander David Bowman and his deputy, Frank Poole, will remain awake. The pair form a minimal caretaker crew during the interplanetary phase of the Saturn mission.

Bowman and Poole are told that their mission aims to expand knowledge of the Solar System and to extend space technology capabilities, and that the survey crew has been placed aboard Discovery in hibernation to conserve life support resources. Hibernation development is a major goal of their mission, for after 100 days of scientific exploration at Saturn the entire human crew is scheduled to hibernate for more than five years. Eventually, the as-yet-unbuilt spacecraft Discovery II will arrive to take them home.

The NASA Glenn Discovery II has no connection with the Discovery II crew-retrieval spacecraft of the novel 2001. Clarke barely describes the latter. I encourage readers to speculate on the shape and capabilities of the Discovery II in the 2001 universe.

The sixth member of the Discovery crew, the HAL 9000 computer, is an artificial intelligence (AI). HAL 9000 knows the true purpose of the trip to Saturn; it is, however, programmed not to tell Bowman and Poole. The secrecy order creates a terrible behavioral conundrum for HAL 9000. Deep in its programming is a directive never to distort information, yet it has been commanded to do just that. This weighs heavily on the advanced AI. HAL 9000 is an innocent being, unable to tamp down what amounts to its conscience. The conflicting directives drive HAL 9000 to neurotic behavior which exacerbates the internal conflict, leading to psychosis and murder between Jupiter and Saturn.

Following the deaths of Poole and the three hibernating crewmen, Bowman is left alone aboard Discovery with HAL 9000. For his own safety, he disconnects the AI. When he finishes, he is the only conscious being within a billion kilometers.

Mission Control belatedly tells Bowman the true purpose of Discovery's mission. He begins a program of study to prepare himself for whatever he will encounter at Saturn. Without HAL 9000 to monitor him in hibernation, his has become a true one-way mission. Increasingly intrigued (and not a little daunted) by the prospect of contact with highly advanced aliens, Bowman is, however, able to put aside thoughts of a lonely death far from home. He even sympathizes with HAL 9000's plight.

The real thing: cylindrical projection of Iapetus image mosaic. The Cassini spacecraft captured images of Iapetus during flybys at different distances and under different lighting conditions; hence some are blurred and others are sharp. The leading hemisphere is at right. Image credit: NASA
Bowman uses Discovery's telescopes to observe the Saturn system. He determines that Iapetus is his goal; the line between the dark and light hemispheres is sharp and obviously artificial. Guided by Earth-based computer control, Discovery of the book successfully fires her plasma jet propulsion system to place herself first into orbit about Saturn and then, with her last drops of hydrogen propellant, into orbit about Iapetus. Bowman then descends in a one-man space pod with the aim of landing atop a giant ("at least a mile high") black monolith standing at the exact center of the white hemisphere of Iapetus. Bowman calls it the Tycho monolith's "big brother."

The big monolith has plans for Bowman; it is, among other things, a Stargate, a space-time shortcut leading into a galactic transit system. He is soon whisked across the Milky Way Galaxy to meet an enigmatic fate. I do not feel qualified to describe the intricacies of Stargate technology, so here I will conclude my overview of the second half of the novel 2001.

Discovery II: this forward view highlights the artificial-gravity section and the spacecraft's lone docking port. Image credit: NASA
Neither did I feel qualified to describe nuclear fusion propulsion technology when I started work on this post, but I think I stand a good chance now of accurately describing the NASA Glenn nuclear-fusion spacecraft Discovery II. If, however, you detect what you believe is an error, it probably is, so please let me know so that I can correct it.                                    

In their documents, the NASA Glenn team describes Jupiter and Saturn versions of its Discovery II. As shown by their estimated weights, only minor differences distinguish the two versions; at 1690-metric-tons, the Jupiter-bound Discovery II would weigh only nine metric tons less than the Saturn-bound version. In keeping with my already-established emphasis on Saturn in this post, I have opted to focus on the Saturn version of NASA Glenn's Discovery II.

The Discovery II design was hatched while an obscure NASA study group called the Decadal Planning Team (DPT) was active. A creation of President William Clinton's Office of Management and Budget, the DPT aimed to articulate a philosophical foundation for NASA advance planning in the 21st century. It did this to prepare the way for new space initiatives during the Presidency of Clinton's Vice President and "space czar," Albert Gore.

"Go anywhere, any time" was an oft-repeated DPT slogan that seems on the face of it to apply well to NASA Glenn's Discovery II. The spacecraft's fusion rocket could in theory propel it to Saturn in 212 days when the planet was at opposition - that is, when Saturn was as close to Earth as it could be. The NASA Glenn team found that, all else being equal, a Saturn voyage at conjunction - that is, when Saturn was on the far side of the Sun and thus as distant from Earth as it could be - would last only 15% longer. Discovery II's course to Saturn in both instances would follow nearly a straight line, not the graceful rising curve of a minimum-energy Hohmann transfer.

At present, nuclear fusion occurs mainly inside stars. Human efforts toward harnessing star power since the 1940s have emphasized fusion bombs. The U.S. exploded the first such weapon in 1952; a nuclear-fission device served as the "spark plug" for triggering the fusion explosion.

Development of electricity-generating fusion reactors, by contrast, has turned out to be more difficult than once assumed. The international ITER project, based in southern France, now hopes to test a prototype commercial fusion reactor in the 2030s.

Earth-based fusion electricity-generation technology would need to advance and considerable additional investigation into almost all engineering aspects of fusion rocketry would be necessary before a fusion engine could become part of NASA's spaceflight tool kit. Nevertheless, the NASA Glenn engineers optimistically predicted that Discovery II's maiden voyage might take place 30 years after they completed their NASA TM - that is, in the year 2035.

Discovery II: this aft view highlights the spheromak fusion reactor, four slush hydrogen tanks, and magnetic rocket nozzle. Image credit: NASA
Discovery II's proposed propulsion fusion reactor would be of a modified Tokamak design. The original Tokamak, a 1970s Russian invention, was a normal torus - the proverbial doughnut - with a magnetically contained high-temperature plasma filling. Discovery II's "spheromak" propulsion reactor, by contrast, would be shaped like a cored apple. Plasma would swirl inside a relatively compact torus with an approximately "D"-shaped cross section. Because it would be compact, it would need less structure and fewer heavy components, such as electromagnets. The spheromak could thus be made much lighter than an equivalent doughnut Tokamak.

Though the NASA Glenn team took pains to make her propulsion reactor as light as possible, at an estimated 310 metric tons it was still the most massive single hardware element of the Discovery II. The reactor weight estimate did not include support systems such as the fission reactor and battery bank that would supply the electrical power necessary for fusion reactor startup.

Tokamak vs. spheromak. Image credit: Culham Centre for Fusion Energy
The NASA Glenn researchers opted for a deuterium/helium-3 (D/He-3) reactor fuel mix, partly because it is relatively well understood and partly because deuterium and helium-3 are relatively plentiful in the outer Solar System. The spacecraft would arrive at Saturn with an empty reactor fuel tank and refuel with deuterium and helium-3 mined from its icy moons and tawny cloud bands. Discovery II would "burn" 11 metric tons of D/He-3 reactor fuel to travel one-way to Saturn.

Nuclear fusion brings together atomic nuclei at high temperatures and pressures. Lightweight nuclei, such as those of various isotopes of helium and hydrogen, yield the most energy per unit, so are generally favored as reactor fuel. When atomic nuclei fuse, they release prodigious energy and create heavier elements. The heavier elements would, over time, build up in Discovery II's fusion plasma, gradually reducing the reactor's performance. In addition, some small portion of the spheromak interior walls would sputter away and mix with the swirling plasma.

Heavy element and wall debris plasma (informally known as "ash") would congregate in a "halo" against the outer wall of the plasma torus through skillful management of interlaced "toroidal" and "poloidal" magnetic fields, then a gutter-like magnetic "divertor" would vent the ash plasma from the aft end of the torus. The vented ash plasma would produce thrust.

The NASA Glenn team proposed increasing that thrust by augmenting the vented ash plasma flow with hydrogen. Contact with ash plasma and passage through a constricted "throat" would heat the hydrogen until it also became plasma. A skeletal "magnetic nozzle" would then expel the plasma mix into space to generate thrust. The divertor and magnetic nozzle would together have a mass of only six metric tons, the NASA Glenn team estimated.

Discovery II would include four cylindrical 37-meter-long propellant tanks containing a total of 861 metric tons of "slush" hydrogen. Chilling the hydrogen until it became slush using an on-board refrigeration system would increase its density, reducing the size and number of hydrogen tanks required.

Two authors of the 2001 and 2005 Discovery II documents - Borowski and Dudzinski - proposed in 1997 a different kind of propulsion plasma augmentation. I described this in the first post of this three-part series (see "More Information" below). Their system had lunar liquid oxygen, or LUNOX, augment hydrogen plasma expelled from a nuclear-fission reactor. The hydrogen plasma and LUNOX would burn as in a chemical-propellant rocket engine, increasing thrust and making possible 24-hour Earth-moon "commuter" flights.

To place its Discovery II fusion ship into space, the NASA Glenn team postulated the existence of a Heavy Lift Launch Vehicle (HLLV) capable of boosting 250 metric tons into a circular assembly orbit between 140 and 260 nautical miles above the Earth. They argued that 250 tons would be very near the practical maximum payload for an HLLV. Placing Discovery II components into assembly orbit would require that seven of the monster rockets launch in rapid succession. This would create challenges in the areas of HLLV assembly, pad installation, and launch operations, among others.

The Discovery spacecraft of the book and film 2001 included large-diameter propulsion and crew modules. The latter was a sphere a little over 12 meters in diameter and the former was even longer and wider. The NASA Glenn team looked upon these with skepticism; such modules would likely be too large to launch intact, so would need to be at least partly built in space by spacewalking astronauts or through complex teleoperations.

They ignored the versatile space pods portrayed in the book and film 2001, which might have made orbital assembly easier, opting instead for pre-assembled launch packages that would fit within a 10-meter-diameter, 37-meter-long streamlined HLLV payload fairing. The self-propelled launch packages would, they explained, rendezvous and dock automatically in assembly orbit.

EASE assembly experiment, 1985. Image credit: NASA
Having said that, they then contradicted themselves by describing a series of HLLV payloads that would in fact require extensive in-space assembly. The first would include Discovery II's 203-meter-long, six-metric-ton central truss. The NASA Glenn team explained that it would be based on the Experimental Assembly of Structures in EVA (EASE) Space Station truss concept tested during STS-61B spacewalks in late 1985.

The Discovery II truss, hexagonal in cross section, would comprise 58 "bays," each built from 97 separate struts, nodes, and other parts. EASE assembly occurred in the payload bay of the Space Shuttle Atlantis. Discovery II central truss assembly would apparently take place in open space.

Following truss assembly, spacewalking astronauts would install a wide variety of systems inside and outside the truss. Most obvious were 20 rectangular 25-meter-long radiator panels for cooling Discovery II's electricity-generation systems. Each would reach orbit folded like an accordion.

Most important for the six remaining HLLV flights of the assembly phase would be communications, avionics, and reaction control systems. The avionics system, linked to controllers on Earth through the communications system, would use the hydrogen-fueled reaction control system to keep the truss and its attached payloads stable in orbit so that subsequent payloads could dock with it.

The HLLV's 250-metric-ton weight limit required that the 310-metric ton reactor reach space in two launches. The NASA Glenn engineers proposed launching part of the fusion reactor - its poloidal magnetic coils - with the truss payload. This meant that spacewalking astronauts would need to piece together in space Discovery II's most complex and important hardware element.

The second HLLV payload would include the remainder of the fusion reactor and the magnetic rocket nozzle. The third - the 172-metric-ton "artificial gravity crew payload" - would comprise seven pressurized crew modules. The NASA Glenn team offered no information as to how the crew modules would join together automatically in assembly orbit. The fourth through seventh payloads would each comprise a filled slush hydrogen propellant "cryo-tank." Thrusters and avionics would permit them to maneuver into place near Discovery II's tail.

Discovery II with selected components and dimensions indicated. Click on image to enlarge. Image credit: NASA
The NASA Glenn team had Discovery II saving reactor fuel and propellant by departing the Earth-moon system from a loose, distant, "sub-parabolic" orbit, but gave no indication as to how she would reach her departure orbit from her assembly orbit. Presumably the spacecraft would be moved using space tugs.

The NASA Glenn engineers stated that an air-breathing space plane would deliver a six-to-12-person crew to a space station in low-Earth orbit. There they would board a taxi vehicle for the journey to the waiting Discovery II. They would dock with Discovery II's only docking port, located on the front of her central hub crew module, and transfer to Discovery II.

Crews returning from Saturn would park the spacecraft in sub-parabolic orbit and await retrieval. Discovery II would be designed for reuse, though how she would be refueled, resupplied, and refurbished in sub-parabolic orbit after each flight was left to the reader's imagination.

Preparation for departure, much of which could occur before the crew arrived, would require weeks. The magnets and reactor structure would need to be thoroughly cooled using liquid helium, though the chief reason for the long preparation period would be the need to charge a five-metric-ton nickel-hydrogen battery bank. A two-megawatt, 10-metric-ton auxiliary fission reactor inside the central truss would slowly charge the batteries in preparation for the roughly one-gigawatt burst of radio-frequency energy needed to start fusion in the reactor. The NASA Glenn team called this start-up technique "high harmonic fast wave heating."

D/He-3 fuel would enter the reactor rather spectacularly as one-gram, 2.2-centimeter cube-shaped "pellets" accelerated at a rate of 27,580 gravities inside a 185-meter-long electromagnetic rail-gun. How the long, complicated rail-gun would be assembled in space within the central truss was not described. The solid-deuterium/liquid-helium-3 pellets would enter the reactor moving at 10 kilometers per second, so would deeply penetrate the dense plasma torus. This would help to preserve the stability of the swirling plasma flow. Pellets would need to be injected into the reactor once per second to maintain reactor energy output.

Swirling plasma in the fusion reactor torus would at start-up torque (twist) the central truss. Discovery II's maximum acceleration would reach 1.9 milligravities as she closed in on her target planet, when her cryo-tanks would be nearly empty. This acceleration, though minute, would place strain on the central truss, as would operation of various turbines and the movement of coolant and working fluids through pipes and pumps. The reactor fuel injector would generate a four-gravity load each time it fired a pellet. The NASA Glenn team suggested that a flywheel might absorb some of the forces Discovery II would place on herself, but provided little information as to how this would function.

The revolving crew section would also place strain on the truss. Though they noted that data concerning a healthful level of artificial gravity do not yet exist, NASA Glenn team opted to provide Discovery II's astronauts with artificial gravity one-fifth as strong as Earth gravity. Three arms 17 meters long would each connect a 5.6-meter-tall, 7.5-meter-diameter lab/hab module to a 7.5-meter-diameter central hub where weightless conditions would prevail. The artificial-gravity system would spin 3.25 times per minute. The two-deck lab/hab modules would contain accommodations for four astronauts each. Opting for separate lab/hab modules connected only through a hub would mean that the "hamster wheel" jogging routine demonstrated in the film 2001 could not occur.

All crew modules would include a layer of water between two layers of graphite epoxy hull material for radiation protection and to serve as a heat dump for crew module thermal control. The central hub would also contain a solar-flare storm shelter with augmented shielding.

Discovery II would arrive at Saturn - more accurately, Saturn sub-parabolic orbit - with nearly empty cryo-tanks. Her crew would carefully shut down her fusion reactor and begin charging her battery bank for another start-up in several weeks' time.

The NASA Glenn team offered a vague vision of how their ship might refuel for the trip home to Earth. Robotic fuel-gathering systems, perhaps suspended from balloons, might be placed into Saturn's atmosphere. They would need to process hundreds of kilograms of gas to obtain a single gram of helium-3 or deuterium and tens of thousands of tons to collect the 11 metric tons required to refuel Discovery II.

No indication was given as to how or when the refueling infrastructure would have been established. Similarly mysterious was how the collected D/He-3 would reach Discovery II.

Hydrogen propellant would be more plentiful than either deuterium or helium-3. The NASA Glenn team envisioned that Discovery II's self-propelled hydrogen cryo-tanks would separate and maneuver to an automated refueling station; meanwhile, identical full tanks would rendezvous with Discovery II and replace the depleted ones. How and when the hydrogen refueling station would have been established and how it would collect hydrogen at Saturn was left to the imagination.

Discovery II would carry no auxiliary craft, so would need vehicles pre-deployed at Saturn if her crew was meant to leave the ship and land on any of the ringed planet's many moons. The NASA Glenn team did not explain how auxiliary craft might reach Saturn ahead of Discovery II, nor how they would be maintained after they became based in Saturn's neighborhood.

Could NASA Glenn's Discovery II replicate the capabilities of 2001's Discovery? As detailed in the second part of this three-part post series, the film Discovery differed from the Discovery of Clarke's novel. The cinematic and literary spacecraft had different propulsion systems, though both relied on nuclear fission. Discovery of the film was a gas-core nuclear-thermal rocket; Discovery of the book employed electromagnetic "plasma jets" that drew electricity from a fission reactor.

Both of 2001's Discovery spacecraft - admittedly fictional, but designed with great concern for realism - could travel round-trip to Jupiter without reliance on pre-deployed assets. Both were adaptable enough that they could be diverted from Jupiter to Saturn when the need arose. That adaptability was based on advanced crew support (hibernation) and automation (HAL 9000) systems. Because of those advanced non-propulsion systems, the round-trip Jupiter mission could be re-planned as a one-way Saturn mission with eventual crew retrieval by a second spacecraft.

Though it promised fusion rocket capability, the NASA Glenn Discovery II design study is in fact incomplete. Among other things, its failure to account for the existence at Saturn of extensive pre-deployed assets essential to Discovery II's mission plan makes it hard to take seriously. Furthermore, it emphasizes propulsion to the exclusion of other potentially ground-breaking, mission-shaping technologies. For these reasons, NASA Glenn's Discovery II cannot be said to replicate the capabilities of the Discovery spacecraft portrayed in the book and film 2001: A Space Odyssey.

Saturn viewed by the Cassini spacecraft from an orbital position north of the planet's equator. The gray north polar region and peculiar spinning hexagonal polar vortex at its center are just visible. Image credit: NASA
More Information

Could the Space Voyages in the Film and Novel "2001: A Space Odyssey" Really Happen? (Part 1)

Could the Space Voyages in the Film and Novel "2001: A Space Odyssey" Really Happen? (Part 2)

The Challenge of the Planets, Part Three: Gravity

Sources (please also see Part 1 and Part 2 Sources)

Realizing "2001: A Space Odyssey": Piloted Spherical Torus Nuclear Fusion Propulsion, NASA/TM-2005-213559, C. Williams, L. Dudzinski, S. Borowski, and A. Juhasz, NASA Glenn Research Center, March 2005

2001: A Space Odyssey, Arthur C. Clarke, New York: New American Library, October 1999, pp. 80-82, 85-101, 120-203