Borderlands of Science

by Charles Sheffield

  The pulsed fission rocket was proposed by Stanislaw Ulam in 1955. The idea, later known as Project Orion, appeared practical and could have been built. However, the effort was abandoned in 1965, a casualty of the 1963 Nuclear Test Ban Treaty. Project Orion called for full-scale atomic explosions, and the treaty made it impossible to test the idea. The EJV is excellent, up to 100 kms/second, but the mass of the pusher plate may limit practical accelerations to a few centimeters/sec2 (less than a hundredth of a gee). This is no good for a launch system, but it will achieve respectable velocities over long periods. An acceleration of 1 cm/sec2 (just over a thousandth of a gee) for one year produces an end speed of 310 kms/second. Note that, even at this speed, Neptune is still more than six months travel time away.

  8.8 Pulsed fusion. The pulsed fission rocket of Project Orion has two big disadvantages. First, the nuclear explosions are full-scale nuclear blasts, each one equivalent in energy release to thousands or even millions of tons of conventional explosives. Second, the massive "pusher plate" is useful as a protection against the blasts and as an absorber of momentum, but it greatly decreases the acceleration of the ship and the system efficiency.

  The pulsed fusion rocket potentially overcomes both these problems. Each fusion explosion can be a small one, involving only a gram or so of matter. The fusion process is initiated by a high-intensity laser or a relativistic electron beam focused on small spheres of nuclear fuel. The resulting inward-traveling shock wave creates temperatures and pressures at which fusion can occur. If the right nuclear fuels are used, all the fusion products can be charged particles. Their subsequent movement can therefore be controlled with electromagnetic fields, so that they do not impinge on the payload or the walls of the drive chamber.

  An analysis of a pulsed fusion rocket mission was performed in the late 1970s by the British Interplanetary Society. Known as Project Daedalus, it was a design for a one-way trip to Barnard's Star, 5.9 light-years from the Sun. Small spheres of deuterium (D) and helium-3 (He3) were used as fusion fuels. (Deuterium is "heavy" hydrogen, 1H2, with a neutron as well as a proton in the nucleus; helium-3, 2He3, is "light" helium, missing a neutron in its nucleus.)

  The D-He3 reaction yields as fusion products a helium nucleus and a proton, both of which carry electric charges and can thus be manipulated by magnetic fields. The estimated EJV for Project Daedalus was 10,000 kms/second, leading to a fifty-year travel time for the 5.9 light-year journey. The mass at launch from solar orbit was 50,000 tons, the final mass was 1,000 tons, and the terminal velocity for the spacecraft was one-eighth of the speed of light.

  The design was a technical tour de force, but the complications and caveats are significant. First, controlled pellet fusion of the type envisaged has never been demonstrated. The D-He3 fusion reaction in the fuel pellets proceeds rapidly only at extreme temperatures, and while other fusion reactions, such as deuterium-tritium, take place at a sixth of this temperature, they produce uncharged neutrons as fusion products and the direction of travel of these uncharged particles cannot easily be controlled.

  Third, and perhaps the biggest problem of all, the nuclear fuels needed are not available. Deuterium is plentiful enough, at one part in 6,000 in ordinary hydrogen. But He3 is very rare on Earth. The total U.S. supply is only a few thousand liters. The Daedalus design calls for 30,000 tons of the stuff, far more than could be found anywhere on Earth. The only place in the solar system where He3 exists in enormous quantities is in the atmospheres of the gas-giant planets, Jupiter and Saturn and Uranus and Neptune.

  Project Daedalus proposed the use of a complicated twenty-year mining operation in the atmosphere of Jupiter, to be conducted by automated factories floating in the Jovian atmosphere. The construction of the spacecraft itself would be carried out near Jupiter. I took over their method in my novel Cold as Ice, but I assumed that the moons of Jupiter had already been colonized by humans. Access—and management oversight—was easier, and in fact the necessary helium mining formed only a minor element of the book.

  8.9 Antimatter rockets. To every particle in nature there corresponds an antiparticle. Matter constructed from these antiparticles is termed antimatter, or mirrormatter. For example, antihydrogen consists of a positron moving about an antiproton, whereas normal hydrogen is an electron moving about a proton.

  When matter and antimatter meet, they annihilate each other. They therefore represent a vast source of potential energy.

  If electrons and positrons meet, the result is high-energy gamma rays, and no particles. If protons and antiprotons meet, the result is an average of three charged pions and two uncharged pions, with the charged pions carrying 60 percent of the total energy. Neutral pions decay to form high-energy gamma rays in less than a thousand trillionth of a second. Charged pions last a lot longer, relatively speaking, decaying to the elementary particle known as a muon in 26 nanoseconds. Muons decay in their turn to electrons and neutrinos, lasting on average 2.2 microseconds before they do so.

  These are short times, but relativity helps here. The charged pions created in this process are traveling fast, at over ninety percent of the speed of light, and thus the effect of relativistic time dilation is to increase their lifetime from 26 nanoseconds to 70 nanoseconds. This is more than long enough to control the movement of the charged pions with magnetic fields. Similarly, the rapidly-moving muons that appear as decay products last on average 6.2 microseconds rather than 2.2 microseconds, before they in turn decay. They too can be controlled through the use of magnetic fields.

  Antimatter is a highly concentrated method of storing energy. The total energy produced by a milligram of antimatter when it meets and annihilates a milligram of ordinary matter is equal to that of twenty tons of liquid hydrogen/LOX fuel. It is therefore ideal for use on interstellar missions, where energy per unit weight is of paramount importance in fuels.

  The most economical way of using such a potent fuel is not to take it "neat," but to dilute the antimatter with a large amount of ordinary matter. Matter/antimatter annihilation then serves to heat up ordinary matter, which is expelled as reaction mass. In this case, both the high-energy gamma rays and the pions serve to heat the reaction mass; and by choosing the antimatter/matter ratio, many different missions can be served with a single engine design. A highly dilute matter/antimatter engine also has excellent potential for interplanetary missions.

  Given all these useful properties of antimatter, why are we waiting? Well, one question remains: How do we get our hands on some of this stuff?

  That leads us to one of the major mysteries of physics and cosmology. There is as much reason for antimatter to exist as for ordinary matter to exist. Logically, the universe should contain equal amounts of each. In practice, however, antimatter is very rarely found in nature. Positrons and antiprotons occur occasionally in cosmic rays, but if we discount the highly unlikely possibility that some of the remote galaxies are all antimatter, then the universe is ordinary matter to an overwhelming extent.

  One product of the recent inflationary models of the early universe is a possible explanation of the reason why there is so little antimatter. This, however, is of little use to us. We need antimatter now, and in substantial quantities, if we are to use matter-antimatter annihilation to take us to the stars.

  Since antimatter is not available in nature, we will have to make our own. And this is possible. One by-product of the big particle accelerators at Fermilab in Illinois, at IHEP in Novosibirsk in the Soviet Union, and at CERN in Switzerland, is a supply of antiprotons and positrons. The antiprotons can be captured, slowed down, and stored in magnetic storage rings. Anti-hydrogen can be produced, by allowing the antiprotons to capture positrons. Antimatter can be stored in electromagnetic ion traps, and safely transported in such containers.

  We are not talking about large quantities of antimatter with today's production methods. Storage rings have held up to a trillion antiprotons, but that is still a very small mass (about a trillionth of a gram). And antimatte
r takes a lot of energy to produce. The energy we will get from the antimatter will not be more than 1/10,000th of the energy that we put into making it. However, the concentrated energy of the end product makes this a unique fuel for propulsion.

  The EJV of a matter/antimatter engine depends on the matter-to-antimatter ratio, and it can be selected to match the needs of particular missions. However, for interstellar travel we can safely assume that we want the biggest value of the EJV that we can get. This will occur when we use a 1:1 ratio of matter to antimatter, and direct the charged pions (and their decay products, the muons) with magnetic control of their final emission direction. Since the charged pions contain 60 percent of the proton-antiproton annihilation energy, and since the uncharged pions and the gamma rays will be emitted in all directions equally, we find the maximum EJV to be 180,000 kms/second. With such an EJV, and a ratio of initial mass to final mass of 3:1, the terminal velocity of the mission will be almost two-thirds of the speed of light. We are in a realm of velocities where relativistic effects have a big effect on shipboard travel times.

  8.10 Photon rockets. This takes the matter-antimatter rocket to its ultimate form. It represents the final word in rocket spaceships that employ known physics.

  If we could completely annihilate matter, so that it appeared as pure radiation (and was heading in the right direction, as a collimated beam), the EJV would be the speed of light, about 300,000 kilometers per second.

  This is the highest EJV possible. It implies perfect magnetic control and redirection of all charged pions, plus the control of all uncharged pions and gamma rays and of all decay products such as electrons and neutrinos. Every particle produced in matter-antimatter annihilation ultimately decays to radiation, or to electrons and positrons that can then annihilate each other to give pure radiation. All this radiation must be emitted in a direction exactly opposite to the spacecraft's motion.

  If the best chemical rocket with a fuel-to-payload ratio of 10,000:1 could be replaced with a photon rocket, the mission would be 99.99 percent payload; the fuel would be a negligible part of the total mass. Having said that, we must also say that we have no idea how to make a photon rocket. It could exist, according to today's physics; but it is quite beyond today's technology.

  8.11 Space travel without reaction mass. The central problem of the rocket spacecraft is easy to identify. For low to moderate EJV's (which we will define as less than 100 kms/second—a value that would make any of today's rocket engineers ecstatic) most of the reaction mass does not go to accelerate the payload. It goes to accelerate the rest of the fuel. This is particularly true in the early stages of the mission, when the rocket may be accelerating a thousand tons of fuel to deliver ten tons of payload. All systems carrying their reaction mass along with them suffer this enormous intrinsic disadvantage. It seems plausible, then, that systems which do not employ reaction mass at all may be the key to successful space travel. We now consider:

  * Gravity swingbys.

  * Solar sails.

  * Laser beam propulsion.

  * The Bussard ramjet.

  Also, we will touch on three hybrid systems:

  * Laser-powered rockets.

  * The Ram Augmented Interstellar Rocket (RAIR).

  * The vacuum energy drive.

  8.12 Gravity swingbys. There is one form of velocity increase that needs neither onboard rockets nor an external propulsion source. In fact, it can hardly be called a propulsion system in the usual sense of the word. If a spacecraft flies close to a planet it can, under the right circumstances, obtain a velocity boost from the planet's gravitational field. This technique is used routinely in interplanetary missions. It was used to get the Galileo spacecraft to Jupiter, and to permit Pioneer 10 and 11 and Voyager 1 and 2 to escape the solar system. Jupiter, with a mass 318 times that of Earth, can give a velocity kick of up to 30 kms/second to a passing spacecraft. So far as the spaceship is concerned, there will be no feeling of onboard acceleration as the speed increases. An observer on the ship experiences free fall, even while accelerating relative to the Sun.

  If onboard fuel is available to produce a velocity change, another type of swingby can do even better. This involves a close approach to the Sun, rather than to one of the planets. The trick is to swoop in close to the solar surface and apply all available thrust near perihelion, the point of closest approach.

  Suppose that your ship has a small velocity far from the Sun. Allow it to drop toward the Sun, so that it comes close enough almost to graze the solar surface. When it is at its closest, use your onboard fuel to give a 10 kms/second kick in speed; then your ship will move away and leave the solar system completely, with a terminal velocity far from the Sun of 110 kms/second.

  The question that inevitably arises with such a boost at perihelion is, where did that "extra" energy come from? If the velocity boost had been given without swooping in close to the Sun, the ship would have left the solar system at 10 kms/second. Simply by arranging that the same boost be given near the Sun, the ship leaves at 110 kms/second. And yet the Sun seems to have done no work. The solar energy has not decreased at all. It sounds impossible, something for nothing.

  The answer to this puzzle is a simple one, but it leaves many people worried. It is based on the fact that kinetic energy changes as the square of velocity, and the argument runs as follows: The Sun increases the speed of the spacecraft during its run towards the solar surface, so that our ship, at rest far from Sol, will be moving at 600 kms/second as it sweeps past the solar photosphere. The kinetic energy of a body with velocity V is V2/2 per unit mass, so for an object moving at 600 kms/second, a 10 kms/second velocity boost increases the kinetic energy per unit mass by (6102-6002)/2=6,050 units. If the same velocity boost had been used to change the speed from 0 to 10 kms/second, the change in kinetic energy per unit mass would have been only 50 units. Thus by applying our speed boost at the right moment, when the velocity is already high, we increase the energy change by a factor of 6,050/50=121, which is equivalent to a factor of 11 (the square root of 121) in final speed. Our 10 kms/second boost has been transformed to a 110 kms/second boost.

  All that the Sun has done to the spaceship is to change the speed relative to the Sun at which the velocity boost is applied. The fact that kinetic energy goes as the square of velocity does the rest.

  If this still seems to be getting something for nothing, in a way it is. Certainly, no penalty is paid for the increased velocity—except for the possible danger of sweeping in so close to the Sun's surface. And the closer that one can come to the center of gravitational attraction when applying a velocity boost, the more gratifying the result.

  Let us push the limits. One cannot go close to the Sun's center without hitting the solar surface, but an approach to within 20 kilometers of the center of a neutron star of solar mass would convert a 10 kms/second velocity boost provided at the right moment to a final departure speed from the neutron star of over 1,500 kms/second. An impressive gain, though the tidal forces derived from a gravitational field of over 10,000,000 gees might leave the ship's passengers a little the worse for wear.

  Suppose one were to perform the swingby with a speed much greater than that obtained by falling from rest? Would the gain in velocity be greater? Unfortunately, it works the other way round. The gain in speed is maximum if you fall in with zero velocity from a long way away. In the case of Sol, the biggest boost you can obtain from your 10 kms/second velocity kick is an extra 100 kms/second. That's not fast enough to take us to Alpha Centauri in a hurry. A speed of 110 kms/second implies a travel time of 11,800 years.

  8.13 Solar sails. If gravity swingbys of the Sun or Jupiter can't take us to the stars fast enough, can anything else? The Sun is a continuous source of a possible propulsive force, namely, solar radiation pressure. Why not build a large sail to accelerate a spacecraft by simple photon and emitted particle pressure?

  We know from our own experience that sunlight pressure is a small force—we don't have to "lean i
nto the sun" to stay upright. Thus a sail of large area will be needed, and since the pressure has to accelerate the sail as well as the payload, we must use a sail of very low mass per unit area.

  The thinnest, lightest sail that we can probably make today is a hexagonal mesh with a mass of about 0.1 grams/square meter. Assuming that the payload masses much less than the sail itself, a ship would accelerate away from Earth orbit to interstellar regions at 0.01 gees.

  This acceleration diminishes farther from the Sun, since radiation pressure per unit area falls off as the inverse square of the distance. Even so, a solar sail starting at 0.01 gees at Earth orbit will be out past Neptune in one year, 5 billion kilometers away from the Sun and traveling at 170 kms/second. Travel time to Alpha Centauri would be 7,500 years. Light pressure from the target star could be used to slow the sail in the second half of the flight.

  8.14 Laser beam propulsion. If the acceleration of a solar sail did not decrease with distance from the Sun, the sail we considered in the last section would have traveled ten times as far in one year, and would be moving at 3,100 kms/second. This prompts the question, can we provide a constant force on a sail, and hence a constant acceleration, by somehow creating a tightly focused beam of radiation that does not fall off with distance?

  Such a focused beam is provided by a laser, and this idea has been explored extensively by Robert Forward in both fact and fiction (see, for example, his Flight of the Dragonfly, aka Rocheworld; Forward, 1990). In his design, a laser beam is generated using the energy of a large solar power satellite near the orbit of Mercury. This is sent to a transmitter lens, hanging stationary out between Saturn and Uranus. This lens is of Fresnel ring type, 1,000 kilometers across, with a mass of 560,000 tons. It can send a laser beam 44 light-years without significant beam spreading, and a circular lightsail with a mass of 80,000 tons and a payload of 3,000 tons can be accelerated at that distance at 0.3 gees. That is enough to move the sail at half the speed of light in 1.6 years.