≡ Menu

Friedwardt Winterberg on Starship Design

Imagine frozen pellets of deuterium and helium-3 being ignited by electron beams to produce fusion, all this occurring in a combustion chamber fully 330 feet in diameter. Such was one early concept for Project Daedalus, the British Interplanetary Society’s starship design that would evolve into a two-stage mission with an engine burn — for each stage — of two years, driving an instrumented payload to Barnard’s Star at twelve percent of the speed of light.

We’ve been kicking the Daedalus concept around here recently because the BIS is developing, in conjunction with the Tau Zero Foundation, Project Icarus, a revisiting of the original Daedalus concept. The Daedalus propulsion system required fifty billion fuel pellets, thirty thousand tons of helium-3 and 20,000 tons of deuterium, as massive an undertaking as our species has ever attempted, given that the helium-3 would have to come from the atmosphere of a gas giant like Jupiter. Icarus will study what Daedalus might look like with newer technologies.


Image: A diagram of the Daedalus starship. Credit: Adrian Mann.

For propulsion inspiration, the original Daedalus team turned to Friedwardt Winterberg, who had studied fusion initiation through electron beams, and it was because of that involvement that Project Icarus team leader Kelvin Long contacted Dr. Winterberg again with news of the Icarus study. Amongst their e-mail exchanges were some comments I found interesting, and because Dr. Winterberg has given permission to use parts of these, I want to run one of them now. In this first excerpt, he speaks of the background of propulsion studies and what may be feasible as we expand into our own Solar System:

When I first had thought of the fusion-micro-explosion propulsion system almost 40 years ago, I never thought about interstellar spaceflight. I rather thought about a high specific impulse – high thrust propulsion system for manned spaceflight within the solar system. Instead of an interstellar probe, one could build in space very large interference “telescopes” with separation distances between the mirrors of 100,000 km, for example, in the hope to get surface details of other earthlike planets. And by going to 500 AU at the location of Einstein gravitational lens –focus, one could use the sun as a telescopic lens with an enormous magnification.

Yes, and it’s fascinating to speculate on how dramatically our observations of other solar systems may change our mission concepts. After all, when Daedalus was envisioned, no one knew whether there were planets around other stars, and Barnard’s Star was chosen specifically because there was at least some evidence of one or more planets there. A flyby probe would be a way to find and characterize these planets, but in fifty years or less, we may be able to see distant exoplanets clearly enough to limit actual missions to specific, high-value targets. Dr. Winterberg continues:

All this needs a very powerful propulsion system. Before going to Alpha Centauri (or Epsilon Eridani), one should aim at comets in the Oort cloud. Since there is water abundantly available, [this] invites the use of deuterium as rocket fuel. Unlike a DT micro-explosion where 80% of the energy goes into neutrons, unsuitable for propulsion, it is not much more than 25% for deuterium. A deuterium mini-detonation though requires at least 100 MJ for ignition, but this can be provided with a magnetically insulated Gigavolt capacitor, driving a 100 MJ proton beam for the ignition of a cylindrical deuterium target…

We’ve looked before at how this might be done, specifically in one of Dr. Winterberg’s papers, as examined by Adam Crowl in this essay. But how do we proceed to put this power to work? A primary destination emerges:

In reaching the Oort cloud, and there establishing human colonies, one may by “hopping” from comet to comet ultimately reach a “new” earth.

Recent research I believe, suggests that in already 100 million years the earth may become inhabitable through the loss of oxygen.

But 100 million years gives us still plenty of time.

Plenty of time indeed. The notion of moving from comet to comet is appealing, invoking as it does the possibility of a gradual expansion deeper and deeper into the Oort, with the prospect of eventually encountering comets in a similar cloud around the Centauri stars. Moving a step at a time may obviate the need for a single interstellar crossing, breaking it into stages. In any case, such a crossing would be an enormous undertaking, as Dr. Winterberg goes on to note:

I cannot see how except with gargantuan space craft interstellar space flight will ever be possible. And I have little taste to speculate about the surprises the physics of tomorrow might bring. It is possible that big surprises still wait for us, but the opposite is possible as well. Discovering the laws of nature is like discovering America. It may happen just once. I therefore like to think what is possible with what we know now.


A sound approach, and one with the added benefit of not requiring new physics to work. The Tau Zero Foundation hopes to encourage both approaches, missions based on known physics and rigorous examination of possible ‘breakthrough’ concepts. We don’t know whether the latter will turn up or not, but we learn valuable lessons about the universe from the attempt to find them even if breakthroughs don’t emerge.

As to the size of a starship, my own guess is that gargantuan craft are not the future for our early interstellar probes, assuming we build such. It may well be that we can couple propulsion advances (possibly via beamed microwave or laser designs) with nanotechnology to produce robotic probes of extremely small size. But assuming we do follow the fusion route, here are Dr. Winterberg’s further thoughts:

And there I think fusion propulsion with deuterium appears quite obvious, with water (in the comets of the Oort cloud) in interstellar space widely available. But as we have known since the paper by Trubnikov (2nd UN Conference in Geneva in 1958 on the peaceful use of atomic energy), deuterium fusion with magnetic confinement is unlikely possible for a fusion reactor of reasonable dimensions. The situation is quite different with a deuterium detonation. There, the DD reaction produces T and He3, which in a secondary reaction burn with D. This was “nicely” demonstrated by the 15 Megaton fission triggered deuterium bomb test in 1952.

For propulsion, the pure fusion fire ball can with much higher efficiency (if compared to a pusher plate) be deflected by a magnetic mirror, also avoiding the ablation of a pusher plate. The ignition, requiring more than 100 MJ, can be done with some kind of particle accelerator. The LHC at CERN can store several 100 MJ energy in a particle beam moving with almost the velocity of light. No laser can that do yet. Unlike lasers, particle accelerators are very efficient. And to get a high fusion yield of say about 1kt, cylindrical targets with axial detonation should be used, where a mega-gauss magnetic field entraps the charged fusion products, as it is required for detonation.

The paper on this work is “Deuterium microbomb rocket propulsion,” which Dr. Winterberg presented at the 2008 Advanced Propulsion Workshop in Pasadena (abstract). Centauri Dreams thanks Dr. Winterberg for being willing to share this correspondence with its audience.

Comments on this entry are closed.

  • Adam April 9, 2009, 18:18

    Hi Paul

    Nice to know I read those earlier papers correctly for their implications :-)

    I agree with him on the potential of the Opik-Oort Cloud (the OOC, sounds almost like a bad soap-opera title.) I was coincidentally pondering the nature of things out in the OOC. Out past about 810 AU in direct sunlight the temperature is below hydrogen’s melting point. But a bit closer in deuterium freezes out before protium does (protium at ~ 14 K and deuterium ~ 18 K.) Some cosmogonic scenarios in favour imply several Earth to Mars sized objects out there and if their early volcanism released hydrogen or they retained their primary atmospheres, then it’s possible that they’ll be covered in seas of protium underlain by deuterium ice. A very handy resource in the OOC.

  • James M. Essig April 10, 2009, 18:20

    Hi Paul;

    The idea of using any form of microfusion device for precursor manned intersteller missions to nearby stars seems more appealing when the microbomb pulse fusion rocket is viewed as an open ended fusion reactor. I would think that a fusion bomblet powered pulsed fusion rocket could be designed and built in relative short order compared to schemes like the interstellar ramjet and any workable improved versions of such yet to be designed.

    The beautiful thing about pulsed fusion rockets is that if one assumes the maximum theoretical Isp for carried along fusion fuel expressed in units of a fraction of C of 0.119 C, then if one assumes that it requires a fueled mass to dry mass fration of 10 or each stage wherein each stage’s dry mass takes the form of a mechanically strong fusion bomblet holder, then the effective maximum fusion fuel Isp is still about 0.108 which can in theory permit an effecive total intitial stage mass to final payload mass of 100,000 to produce a terminal cruise velocity of about 0.86 C or a gamma factor of two assumming perfect system efficienccy which in reality will not be obtainable.

    However, gamma factors of 1.3 to 1.5 might reasonably be obtained for such ratios with optimal bomblet types assumming that the payload includes electrodynamically reactive field effect drag inducing means such as a magnetic breaking coil and the like thus avoiding the need for slow down reverse thrust fuel.

    With medical life span enhancement to indefinate lenghts, the cosmos seems wide open for us to explore under nuclear power. I cannot think of a better use for pure fusion microbombs.

  • Adam April 10, 2009, 19:42

    Hi Paul (again)

    Where does he pull the 100 million year figure from? Sounds like James Lovelock’s or Mike Hart’s old predictions about the biosphere’s time-limits, but that has been superseded by Kasting and Caldeira’s work. Still there’s a few lines of evidence that without planetary engineering on a grand scale Earth will be a dry desert in 0.5-1.0 billion years. From what I’ve read over the years the oceans will drain into the mantle in ~ 500 million years, roughly the same time that C-3 plants go extinct from carbon dioxide draw-down as the Earth warms up. Plate tectonics will seize up at the same time according to some estimates and the magnetic field will run out of puff too.

    Hmmm… perhaps (natural) planets aren’t such a good choice of habitat. They may not be as long-term stable as imagined. I wonder what we can engineer that will be better? Paul Birch’s Supra-mundane planets might be preferable to O’Neill CylCits – an open sky versus a concrete floor. A Supra-Jupiter at the 1 gee level would be ~336 times Earth’s area in size. The gas giants all up would provide about ~500 times Earth’s area in habitat and have the advantage of natural, long-lived magnetic fields.

    Ultimately mining the Sun for heavy elements would provide ~20 Jupiter masses of construction materials and allow the Sun to be engineered for long-life as Martin Beech has described. Starlifting most of its unfused hydrogen would provide enough mass to create several new low mass stars, which might also be an astroengineering challenge worthy of our descendents over the next ~50 billion years.

    But, as already noted, Winterberg’s OOC migration might make engineering the Sun kind of irrelevant to Life at large. Half the Sun’s mass will be returned to the Galaxy and low-temperature post-biological Life might be more interested in the large-scale evolution of all the stars, not just our single star.

    • Administrator April 11, 2009, 8:45

      Adam writes:

      Hmmm… perhaps (natural) planets aren’t such a good choice of habitat. They may not be as long-term stable as imagined. I wonder what we can engineer that will be better? Paul Birch’s Supra-mundane planets might be preferable to O’Neill CylCits – an open sky versus a concrete floor. A Supra-Jupiter at the 1 gee level would be ~336 times Earth’s area in size. The gas giants all up would provide about ~500 times Earth’s area in habitat and have the advantage of natural, long-lived magnetic fields.

      Fascinating to speculate. Iain Banks’ Masaq orbital (from Look to Windward, which I’ve just read, comes immediately to mind, and that brings back Larry Niven’s Ringworld and, of course, all the Dyson thinking. Do you have a reference for Paul Birch’s ideas — I’d like to read up on them.

  • Adam April 12, 2009, 3:52

    Hi Paul

    All of Paul Birch’s refereed papers were published in the JBIS, plus “A Visit to Supra-Jupiter” in “Analog”, and he has them available online at his webpage…


    …he had a falling out with the BIS and one of his requested conference papers was never published and so after several broken promises, much to his disgust, so he hasn’t had dealings with them since. The text of the paper can be found here…


    …it’s is a mind-blower as he discusses a very quick way of cooling Venus, plus the ultimate artificial Planet, which is a gargantuan 1.2 light-years in radius with 30 million levels. Its mass-energy is so high that the inner shell experiences a time dilation of 2500 (!) relative to the wider Universe. A worthy task for a civilization with a billion years to spare.

  • Benjamin April 12, 2009, 4:00

    I find the idea of interstellar bridges of icy and/or rocky bodies stretching between stars fascinating. There’s something terribly appealing in a science-fiction sense of dark frontiers littered with jewels of fusion fuel and minerals; however I do have a concern about sustainability.

    After you have used the space between, say, Sol and Alpha Centauri, for some period of time, how can we avoid irreversibly depleting the available, useful energy supply and forever locking our descendants in our own solar system, or indeed in their target systems?

  • Adam April 12, 2009, 21:53

    Hi Benjamin

    Ultimately large scale trade between systems will require energy beams of some kind powered by solar energy. Electromagnetic accelerator/deccelerator systems could, theoretically, operate with very low overall energy losses because incoming cargo returns energy to the system by regenerative braking. A continuous loop of transport pods could be developed to provide rapid transit between the stars eventually.

    But that’s all a long way off. To build such things requires massive efforts in the Oort Clouds of the stars involved and to trek around the Clouds will require Winterberg’s fusion rockets or some such.

  • Hucbald April 13, 2009, 19:31

    Can we PLEASE ban any and all spacecraft names like Daedalus and Icarus – anything that ends in “-us”? It’s so hopelessly twentieth-century. Besides, Icarus’ wax wings melted when he flew too close to the sun, and he fell in a burned up. That’s a truly knot-headed name for a starship, don’t you think?

    My vote would be for Commerce, or something along those lines. ;^)

  • Administrator April 13, 2009, 19:37


    Good grief. My vote is firmly with Greek mythology!

  • dennymack April 13, 2009, 21:01

    It would seem that the realizable step would be to build a single stage fusion powered booster to send a probe on a flyby. Our imaging and information gathering systems are so much better than they were decades ago that it would seem we can get a lot of bang for buck by skipping the decel side of the trip. Build a redundant drone and let it shoot past Barnards and give us the images and whatever measurements can be taken remotely.
    We might even be able to double dip if we gave the probe a second stage that instead of decel just gave it lateral acceleration to line up on the next star. (I’m not an astronomer, so I don’t know what is “behind” Barnard’s star.)
    If these rigs are huge and expensive, would it be possible to have multiple missions launched on the same booster? Rather than a single second stage, set up several second stages that, once accelerated towards a quadrant of space, would set their individual courses by accelerating away from the original line. That way, if you got funding for the mission once, there would be several periods of boosterism for exploration as we awaited capture of new data packages from probes. It would be a sad thing if we poured all our treasure into one shot and got a flyby of a dull rock. Sure, it would be neat, but we need a way to get the public payer onboard.

  • Howard T. April 14, 2009, 8:57

    Mr. Essig….

    What is a ‘magnetic breaking coil’?

    Are you thinking of ‘braking’ by use of magnetic tethers?

    Where does a coil come in?

    • Administrator April 14, 2009, 9:47

      One place to learn more about superconducting braking coils for interstellar missions is a report that Robert Zubrin and Andrew Martin did on this technology for NIAC back in 2000:


      Zubrin and Dana Andrews were pioneers in looking at magsail possibilities for deceleration, and in this report Zubrin and Martin consider how they could be used for fast missions within the Solar System as well. They go into the details of using these superconducting wire coils for various missions and analyze coil characteristics under various configurations.

      Here’s a bit of the report:

      “A loop of superconducting cable perhaps tens of kilometers in diameter is stored on a drum attached to a payload spacecraft. When the time comes for operation, the cable is played out and a current is initiated in the loop. This current once initiated, will be maintained indefinitely in the superconductor without further power. The magnetic field created by the current will impart a hoop stress to the loop aiding the deployment and eventually forcing it to a rigid circular shape. The loop operates at low field strengths, typically 0.0001 Tesla, so little structural strengthening is required. The loop can be positioned with its dipole axis at any angle with respect to the plasma wind, with the two extreme cases examined for analytical purposes being the axial configuration, in which the dipole axis is parallel to the wind, and the normal configuration, in which the dipole axis is perpendicular to the wind.”

      And so on. Diagrams are furnished in the report.

  • Adam April 15, 2009, 18:08

    Hi dennymack

    Before we launch probes for a closer look we’ll have a good idea about the planets in the target systems. In-space optical instruments will make most of the work of flyby missions quite pointless and only rendevous missions will really produce sufficient data return for the effort expended. Unless flybys are small, cheap and fast in which case we can fire off lots of them.

  • Howard T. May 7, 2009, 10:00


    I’d like ro read some, if not all, of the essays, but my blackberry cannot handle the black on dark grey format.

    Is there any way around this?

  • Howard T. May 8, 2009, 14:45

    There’s something about using fusion bombs for interstellar propulsion that just bothers me…. Won’t they leave an environment littered with radioactice debris behind? An obstacle to any further voyages.

    In any case, haven’t we already graduated to being able to generate antimatter propulsion? And won’t this be completely free of debris, as well as being much more efficient?

    Maybe we could actually send a probe thusly powered… And brake it with magnetic coils.

    Somehow the sail concept seems way too fragile. It would greatly multiply the frontal area and increase the likelihood of damage from collisions.

    How to avoid damage from dust/pebbles, etc. at .1c may be the stopper problem. You can’t dodge around like the Millenium Falcon, even if you could somehow detect oncoming objects with radar.

  • JD May 19, 2009, 11:58


    I have a question about this proposal of Dr. Winterberg.

    To build up these currents, wouldn’t you have to drag a huge nuclear power reactor along, effectively destroying your mass ratio? After all controlled net gain fusion hasn’t been achieved yet.

    Isn’t it that unless we achieve net gain fusion, pure fusion rockets remain a fantasy?