≡ Menu

Toward a Beamed Core Drive

If you didn’t see this morning’s spectacular launch of the SpaceX Falcon 9, be sure to check out the video (and it would be a good day to follow @elonmusk on Twitter, too). As we open the era of private launches to resupply the International Space Station, it’s humbling to contrast how exhilarating this morning feels with the great distances we have to traverse before missions to another star become a serious possibility. We’ve been talking the last few days about the promise of antimatter, but while the potential for liberating massive amounts of energy is undeniable, the problems of achieving antimatter propulsion are huge.

So we have to make a lot of leaps when speculating about what might happen. But let’s assume just for the sake of argument that the problem analyzed yesterday — how to produce antimatter in quantity — is solved. What kind of antimatter engine would we build? If everything else were optimum, we’d surely try to master a beamed core drive, the pure product of the matter/antimatter annihilation sequence. Protons and antiprotons are injected into a magnetic nozzle, blowing out the back at a substantial percentage of the speed of light. This is the kind of rocket analyzed by Ronan Keane (Western Reserve Academy) and Wei-Ming Zhang (Kent State University) in the paper I’ve been skirting around the edges of these past few days.

Channeling Antimatter’s Energies

The paper, headed for publication in the Journal of the British Interplanetary Society, has the provocative title “Beamed Core Antimatter Propulsion: Engine Design and Optimization,” and it deals with the particle stream emerging from proton/antiproton collisions. What you get when you put the two together are gamma rays and pions, some of the latter charged and some neutral. Almost immediately the pions decay into positrons and electrons, which meet each other and produce gamma rays. But the tens of nanoseconds the pions take to decay gives us long enough to channel the charged pions through a magnetic nozzle to produce the needed thrust.

Image: Antimatter promises fast transportation throughout the Solar System and the opportunity for interstellar probes, but only if we can master its production and storage. New work is explaining how efficient an antimatter engine might be. Credit: Positronics Research, LLC.

The beamed core engine, then, is all about channeling the pions into a focused flow. Get this right and you’ve got a lot of energy to work with. In fact, Keane and Zhang note that the energy released per kilogram of annihilating antimatter and matter is 9 X 1016 joules, which is two billion times more than the thermal energy from burning a kilogram of hydrocarbon, and over a thousand times larger than burning a kilogram of fuel in a nuclear fission reactor. But while the beamed core engine is attractive because of the high relativistic velocities of the charged particles produced by the annihilation reactions, the situation is not ideal.

For one thing, much of the energy of the reaction goes into producing electrically neutral particles, which are impervious to the workings of a magnetic nozzle and thus cannot contribute to thrust. The other problem is that the nozzles we’ve been able to analyze have efficiency problems of their own in terms of creating the tight beam of thrust we’d like to produce. What Keane and Zhang do is to use software called Geant4 from the CERN accelerator laboratory to produce simulations of the interactions of particles with matter and fields. They want to bring previous studies of beamed core concepts up to date especially in terms of magnetic nozzles.

Robert Frisbee has performed rigorous studies of beamed core concepts in which magnetic nozzle efficiency is only about 36 percent, which means that while you’re dealing with pions that are initially moving at 90 percent of light speed and above, the exhaust velocity of the rocket would be just a third of that amount. Keane and Zhang derive an efficiency that is better than twice that, and manage to reach charged pion exhaust speeds of 69 percent of c. They also show that the initial speed of charged pions in a beamed core engine is actually closer to .81c than Frisbee’s 90 percent-plus. Despite the lower initial speed, the nozzle efficiencies make quite a difference depending on the kind of mission being attempted:

Frisbee’s papers explain in depth the needed generalization to account for emission of uncharged particles… When loss of propellant is taken into account, Frisbee has shown that ve ~ 0.3c leads to a beamed core rocket facing daunting challenges in reaching a true relativistic cruise speed on a one-way interstellar mission where deceleration at the destination (a “rendezvous” mission) would be involved.

Fuel requirements become critical with lower nozzle efficiencies:

… with a payload of 100 metric tons, a 4-stage beamed core rocket designed for a cruise speed of 0.42c on a 40 light-year rendezvous mission would require 40 million tons of antimatter fuel. If the cruise speed were limited to 0.25c or less, only two stages might be needed, and Frisbee envisaged viable interstellar missions with as few as one beamed core stage; in such scenarios, fuel requirements would be dramatically lower.

All of this at 36 percent nozzle efficiency. The new numbers change the picture, with Keane and Zhang stating “With the new reference point of ve =0.69c provided by the present Geant-based simulation, true relativistic speeds once more become a possibility using the highest performance beamed core propulsion in the distant future.”

Note the ‘distant future’ caveat, highly significant when you consider our problems in producing antimatter (or harvesting same) and the perhaps even more intractable issues involved in storage.

On Software and Methodology

But even if we can’t put a timeframe on something as futuristic as a beamed core rocket, we can continue to study the concept, and it’s heartening to see Keane and Zhang’s conclusion that the simulation software at CERN has proven robust in meeting this challenge and updating our numbers. Whether or not Keane and Zhang’s methodology is on target may be another issue, as Adam Crowl noted recently in a post to a private mailing list of aerospace engineers. Crowl hastens to add that his computations are provisional, but let me quote (with his permission) where he is right now on the magnetic nozzle efficiency issue:

There’s a problem with using just the exhaust velocity given to *part* of the fuel/propellant. It means the actual mass-ratio for a given delta-vee is quite different to a naive computation using the classic Tsiolkovskii equation. A more useful figure of merit for rockets with mass-loss in addition to reaction mass is specific impulse – momentum change per unit mass of fuel/propellant. Using the equations derived by Shawn Westmoreland and the rather vague particle energies in Zhang & Keane, the effective specific impulse is ~0.28c. Even with a perfect jet efficiency the Isp is just 0.31c.

The antimatter reaction, then, may not offer as much as we hoped:

The 0.81c average particle speed quoted in the paper isn’t as useful as the spread of kinetic energies in the particles produced, or the total kinetic energy in the distribution, but they don’t report either figure. What it does imply is that an antimatter-matter reaction puts about 11% of the mass-energy into the charged particles. Not exactly spectacular.

The chance to go to work on concepts through papers in the preprint process is invaluable, and we’ll see how Crowl’s thinking, as well as Keane and Zhang’s, evolves with further study of the issues in this paper. One thing is for sure: Given the manifest problems of antimatter production and storage, we’ll have no shortage of time in which to consider these matters before the question of actually producing this kind of antimatter rocket becomes pressing.

The paper is Keane and Zhang, “Beamed Core Antimatter Propulsion: Engine Design and Optimization,” accepted by JBIS (preprint).


Comments on this entry are closed.

  • Tony P May 22, 2012, 11:39

    Interesting as always. Thanks Paul.

    Question: Has any research been done on creating antimatter on the fly?

    I heard about NASA scientists discovering that thunderstorms can create antimatter in our atmosphere in very limited quantities. I wonder if a method could be devised to replicate and make the process much more efficient. Just a thought.

    Source: http://www.nasa.gov/mission_pages/GLAST/news/fermi-thunderstorms.html

  • tom baty May 22, 2012, 16:35

    Tony, excellant point you made, now you have me thinking along the same lines.Thanks for the link. The beam core maybe the ultimate way to go, but it seems that a rocket that uses a working fluid as a propellant which is heated up by AM would be at first a better way to go. It would be simplier (comparatively) and if you used H2O as the working fluid/propellant you could refuel almost anywhere in the solar system. Just my $.02 adjusted for inflation.

  • tom baty May 22, 2012, 16:42

    PS. Just another thought, if you can make AM on the fly, per Tony’s comment, that would resolve the storage issue. That’s a big plus in itself.

  • David May 22, 2012, 21:54

    I saw a story that said SpaceX thinks it can get a human to Mars in 15 years.
    I am sorry I could not find it but you may enjoy this espcially since the commentators diverge off into Colony Ships


  • Alex Tolley May 22, 2012, 23:24

    Peter Watts’ Blindsight may have a similar concept – beam the AM to the spacecraft. In this case it was “teleported” rather than a directed beam.

  • Eniac May 23, 2012, 7:09

    Tony P and tom baty: Antimatter on the fly is a non-starter, because you will need to put at least as much energy in as you get out.

  • Tim May 23, 2012, 8:26

    Is there any reason this research only applies to the magnetic confinement of annihilation products, and not any charged particles? Could this have implications for other high temperature drives, such as a fusion rocket?

  • tom baty May 23, 2012, 12:48

    Eniac….what you say is very true, I agree. But this applies to any system. There is no such thing as a free lunch, you are changing one form of energy into another and the best is the one which is most effective and efficent. In the case of AM perhaps solar energy could be used, with a reactor you might just as well use that as a direct energy source and save weight. If indeed a way to store AM is found maybe solar energy could be used along with getting AM directly from the atmosphere (per Tony’s link). A refueling station could then be used in earth orbit. Lots of major ands, ifs and buts.

  • Bounty May 24, 2012, 20:09

    A directed beam of (charged?) AM seems interesting. Fixes part of your weight problem. Hey, while they’re sending AM, maybe they could send up some M to react with.

  • kittlej May 24, 2012, 22:21

    a method for large scale energy storage
    Despite its promise antimatter is difficult to make store or utilize with any efficiency
    The only really low mass, readily usable energy that I have come across is involves creating a very large,, high density magnetic field . The energy content of such a field is proportional to its volume and proportional to the square of the field strength. I and imagine a large loop of superconducting material rotating to match the lorentz force the tends to contract the loop ( where the axis of loop rotation is perpendicular to the plane of the loop, and in same as the field direction as the field as it passes though the center of the loop, thus the magnetic field itself does not rotate. )
    The energy stored is thus proportional the the cube of the loop diameter while the mass of the divide is ( roughly ) proportional to to the square of the loop diameter. Adding or harvesting energy to the loop involves increasing ( or utilizing) the current in the loop ( energy proportional to the square of the the current, and linear with the loop circumference) This equation allows for HUGE amounts of energy storage, at a low weight., With a superconductor that operates at a temperature of 50K, the need for coolant is minimized in the outer solar system, or with adequate solar shielding in the inner solar system.
    How would such a system work? Energy would be generated ( nuclear,) or collected – (solar) at a location in the inner solar system and transmitted to the storage loop.The storage loop is not moving in relation to the( much more massive) energy source and parked less than a few hundred kilometers away. The loop is charged up at a rate of a megawatt or so of power for a period of a few years. Given that there is no appreciable loss in the loop this same energy can be withdrawn over a sustained period for years! the only increase in system mass is minimal ( E=MCsquared_)
    How can such energy be used ? for moderate final speed we are looking at highly accelerated neutral atoms ( 10 to 20% of C) moving parallel to the field, or if the storage is really efficient, the use of light pressure from a laser source.
    The magnetic field itself can be used to accelerate paramagnetic atoms ( like atomic hydrogen with its unpaired electron, or the paramagnetic oxygen molecule ( O2) . Paramagnetic acceleration is a separate topic.
    On a per weight basis , a properly scaled magnetic storage loop in free space is going to be relatively easy to understand, easy to charge and may be able to store at least as much energy as a problematic ” tank” of antimatter. In any case converting the energy ( electricity from the loop) into propulsion is far easier and probably safer than an antimatter system. magnetic energy strorage using superconducting loops is an area of active development for practical applications here on earth, but may actually be easier ( ultimately) to deploy in space where the size of the loop and the intensity of the field is not such and issue.
    Think about it and I invite comments,


  • Brett Bellmore May 25, 2012, 14:11

    “I and imagine a large loop of superconducting material rotating to match the lorentz force the tends to contract the loop ”

    Problem here is that you’ve got the sign of the force backwards. Parallel currents attract, opposing currents repel: A current loop is subject to radial forces which try to expand it, not contract it.

  • Eniac May 25, 2012, 23:48

    kittlej: Brett is correct. If you do the math, you will see that the energy stored in a loop is limited by two things: It’s tensile strength in the face of radial forces (which are outwards, as Brett says and contrary to your assertion), and the maximum current density in the superconductor. As fate wills, the result comes out saying that the energy per mass of conductor is right up there with a tank of gasoline, or tanks of H2-O2, or a flywheel. All in the same ballpark, because ultimately they all rely on the chemical bonding of atoms.

    Only nuclear energy of some sort can best this energy density. By many orders of magnitude.

  • Seth May 27, 2012, 18:54

    How would you keep the rocket from becoming a cloud of million-degree plasma? The amount of waste heat your thermal radiators would need to reject would probably make them glow white-hot.

  • Eniac May 28, 2012, 0:03


    How would you keep the rocket from becoming a cloud of million-degree plasma? The amount of waste heat your thermal radiators would need to reject would probably make them glow white-hot.

    I suppose you would try to insulate the actual million degree plasma from the rocket itself so that only a minute fraction of its waste energy is transferred from the burn area to the rocket, the rest goes out the back before it can cause any harm. This is how chemical rockets work, and for nuclear rockets, including antimatter, the principle is even more essential. That is one reason why nuclear-electric ion drive will never be very effective, although it may have its niche in interplanetary propulsion.