Building Blocks for a Generation Ship

posted by Andreas Hein on April 2, 2014

generation ship

Project Hyperion is working on the first ever design of a manned interstellar vessel. Recently, the team thought about how to leverage on existing heritage for the spacecraft’s subsystems. In particular, two major elements of a crewed spacecraft were of particular interest: the propulsion system and the habitat. A preliminary analysis has already been done by the team about two years ago [1].

In a first step, all the options for the propulsion system and the habitat were enumerated. The following criteria were used to select a particular option:

- Level of detail of heritage design

- Applicability to a manned interstellar mission

- Credibility of the heritage design

Options for the propulsion system included Daedalus, Project Icarus’ Ghost Ship, as well as laser sails [2, 3]. For the habitat, various O’Neill colony designs, the Stanford Torus, as well as the 1976 MIT space colony design and the National Space Society’s Kalpana One colony were considered [4-6]. Based on these criteria, the current Project Icarus, Ghost Team’s design was chosen for the propulsion system and the Stanford Torus design for the habitat. The Stanford Torus is the result of a 1975 NASA Summer Study, conducted under participation of such illustrious personalities as Gerard O’Neill and Eric Drexler [5]. The Torus has the capability to sustain the lives of 10,000 to 14,000 inhabitants. This number is in accordance with the recent results obtained by Cameron Smith for the minimum size of a generation ship’s population [7]. However, in order to make the habitat “interstellar-ready”, various modifications are necessary:

- The colony cannot be illuminated by Sunlight and has to be modified to an internal light source

- Radiation shielding has to be modified in order to shield against the increased galactic cosmic rays (GCRs) in interstellar space; extremely energetic particles, which have their origin in violent cosmic events, such as supernovae

- Shielding against interstellar matter has to be provided

- An internal power source is needed, as Sunlight is not available in interstellar space

The Ghost Team’s design has to be modified as well:

- Due to the vastly increased payload (the science payload of the Ghost Ship was 150 tonnes), much more fuel has to be carried on board

- Magsail deceleration is probably not optimal for lower traveling speeds of ~2%c

The team is currently dealing with the following design decisions:

- Radiation shielding options, based on using fusion propellant

- Propulsion system configurations, such as bundling several engines

- Developing a simulation for the mission sequence

- Assess the depth of modifications necessary for the propulsion system and habitat

The Project Hyperion team will present its results of this heritage-based study in the course of 2014. A first result of a high-level analysis can be seen in the image at the top of this article.


[1] Hein, A. M., Pak, M., Pütz, D., Bühler, C., & Reiss, P. (2012). World Ships—Architectures & Feasibility RevisitedJournal of the British Interplanetary Society65(4), 119.

[2] Bond, A., & Martin, A. R. (1978). Project Daedalus. Journal of the British Interplanetary Society31, 5.


[4] O’Neill, G. K. (1977). The High Frontier Human Colonies in Space, William Morrow and Company. Inc., New York.

[5] Johnson, Holbrow (1977). “Space Settlements: A Design Study”. National Aeronautics and Space Administration.

[6] Bajoria, A., Arora, N., & Globus, A. (2006). Kalpana One: A New Space Colony Design. In Earth & Space 2006@ sEngineering, Construction, and Operations in Challenging Environment (pp. 1-8). ASCE.

[7] Smith, C. M. (2014). Estimation of a genetically viable population for multigenerational interstellar voyaging: Review and data for project Hyperion.Acta Astronautica97, 16-29.

Be Sociable, Share!

26 Responses to Building Blocks for a Generation Ship

  1. Antonio says:


    the [1] link doesn’t work.

  2. Andreas Hein says:

    The link should work now.

  3. A very inspiring design. I would be a bit worried about the radiation from the drive, though. Have you checked the radiation flux at the wheel? And although I’m certain the illustration is just a first pitch, I wonder if there are not too many spokes. I expect the wheel would be a tension structure, completely self contained structurally, and that the spokes would be used just to transmit the thrust and to keep the wheel centered. Just a few spokes would actually be used for transportation. So you could reduce the number of spokes; it would give the ship a bit less of a bicycle look ;-)

    • would have a spiral pattern for the spoke; the design would have a cable suspension system much like the Golden Gate Bridge, this would translate torsion forces from star ship rotation from the torus to the center.
      perhaps only two or three cables would need as they spiral out from the spoke with a great many suspenders as augmentation,these also would not be a vertical system as we are in zero gee with centrifugal forces

      • Hi Steven,

        Once the torus is spinning, there will be practically no torsion from the hub. And if the torus is rocket accelerated, then there is no real torsion even at spin up.
        Some kind of torsion member is required, but it can be very light, just a cable as you mention. Spiral spokes are longer and are therefore heavier.
        The torus doesn’t need to be suspended from the center; imagine you have a suspension bridge and you pull up the ends, and turn it into a ring, the big cables will merge into the deck, and the suspenders will entirely disappear. In a spinning torus, the cables are built into the deck. The spokes are mostly for transportation; they do also transmit the thrust from the drive to the torus, but this is a very small force, since the drive has only about 100 tons of thrust. The torus will probably have something like 10 000 tons of centripetal force or more, so the deck will need to be strong. Have you read the novel Ringworld? That’s just about the best illustration of the structural design of this that you can find.

        • John Pattullo says:

          remember though when the ship is under acceleration there will need to be cable tied forward of the hub (assuming this would be the lightest method of pplying the acceleration to the torus) and that will create a inward force on the torus – depends on acceleration but could end up with anything from no force on torus to compression or tension forces

          • absolutely right. The acceleration is pretty weak though. The cable you suggest will be under tension, the spinning torus under much higher tension. There will be some compression mainly in the vertical structure between the drive and the torus, but also a lot of varied forces from all those tanks.

  4. I also use this system to keep generation starships made out of hollowed out asteroids together

  5. It would also be interesting to see the ghost ship combined with the habitat design of Steve Summerford, which I feel is a more modern design than the Stanford Torus.

    • Andreas Hein says:

      Michel, Steve’s habitat design was done from an urban architecture perspective. It is innovative in proposing alternative habitat structures. The Stanford Torus is an integrated study which takes various aspects of a habitat into account, including all the engineering analysis. It is currently the most mature habitat design to date.

      • Ah yes, I see what you mean. I guess Steve’s ideas could be integrated into the torus, since the original study did not say much about interior arrangements. Six segments with bulkheads between them would be an interesting possibility.

  6. Gerry says:

    As a precursor to manned interstellar trips, a version of the Ghost-Ship engine combined with a modest habitat ring might make a great interplanetary vessel too, with a crew of a hundred or more, given the high thrust and a delta v of >.02 c. Perhaps a 5-10 year manned expedition to the worldlets of the Kuiper Belt?

    • John Pattullo says:

      is there really much humans could do out there that couldn’t be done cheaper and better by a probe?

      better to use it for trips to mars or europa – overkill maybe but would still get experiance flying such a ship

      • Hi John,
        Things can be done cheaper by probes, but not really better. Specially not at light days of distance. But anyway, the real point of sending a crewed probe out would be the trip itself, not the destination.
        Icarus or Hyperion type probes are ridiculously overpowered for solar system travel. They would probably reach Jupiter in a month or so.

        • John Pattullo says:

          yeah but frankly the oort cloud isn’t that intresting and very very dark – better to fly to jupiter and back etc several times – hell if nothing else bet rich tourists would pay a shed load for a low orbit over jupiters red spot

          • Jupiter would be a great destination, wouldn’t it? But again, the Hyperion is a very specialized craft. More like a moving suburb than a cruise ship. When it eventually ‘flies’, there should be plenty of other smaller ships around to do that kind of trip. Perhaps the best analogy would be a nuclear submarine, compared to a cruise ship. Yes, you could visit the Bahamas on a nuclear submarine, but it makes a lot more sense to go an a cruise ship ;-)

        • Gerry says:

          I dunno…a one month trip to Jupiter might in fact be an attractive goal when considering manned expeditions and their attendant deep-space-related health hazards.

          I guess it might be a question of which means of providing safety and comfort to human explorers is cheaper and more practical: a super-powered ultra fast ship on the one hand, or a slower ship designed with better radiation shielding, artificial-g, self-contained environmental loops, etc. on the other.

          Time will tell, I hope!

          • Hi Gerry,

            To really answer this, I needed a travel time calculator, you can find it here:, copy it and use it to your heart’s content, if you like. It is very rough, but I think it gives the right order of magnitude. To do different scenarios, just copy columns. Icarus could reach Jupiter in a month, Hyperion would take about 150 days, and a little bit more than a year for the Kuiper belt. A really powerful torch ship could reach and orbit jupiter in a week, accelerating and decelerating at 1g. There are huge gains to going slowly; just changing the trip time from 30 days to 150 days allows you to go from a 5000 ton ship to a 100 000 tons ship for about the same fuel usage.
            Hope you find this useful.

  7. Asher Syed says:


    Just one question. Is this model mechanically stable when accelerating near 0.10c speed?

    Am I missing something?

    Best regards

    • Hi Asher,
      It is probably not stable in a static way. but it doesn’t need to be, it can be continuously dynamically adjusted. Have you ever seem someone balancing a spinning plate on a stick, and then balancing the stick on their chin? That’s pretty much what this design will need to do. Youtube has a bunch of examples. Another example is modern fighter planes; they are all unstable and require constant computer adjustment just to fly level.
      Just a detail , the ship will travel at about 0.01C, or 1% of c. Regards, Michel L

      • Asher Syed says:

        Hi Michel,

        Thank you for replying. I think the wheel will be attached with the central tube via steel wires or similar. If the ship is accelerating in the direction of the tube and the wheel is also rotating to give 9.8 m/s2 rotational acceleration the wires may not withstand the tension.



        • Hi Asher,

          The final velocity is high but the ship’s acceleration will be very gentle. The average acceleration of Icarus is 0,02m/s2; I expect a ship like this would accelerate even more gently. But for many many years.

  8. John Pattullo says:

    quick thought on the radiators, why not use them as at least part of the radiation shield? hell the shield would need to radiate away any heat it absorbed anyway, need radiators to rid the craft of heat from the engines 2 for 1 deal? yes? no?

    • Andreas Hein says:

      Hi John, the requirements for materials which can be used for radiators is not ideal for shielding radiation. Radiators are best if both sides are facing free space. Using them as rad shielding will reduce the radiator’s effective area considerably. There are other factors to consider as well. Thus, it is a complex trade-off.

  9. An important first step that is needed, before a generation ship is built is to master Artificial Closed Ecosystems (ACEs) on the Earth’s surface, where it is relatively safe. This can be achieved in three logical stages:
    1. Develop the ACE-Certified (zero-pollution) infrastructure needed for the inside of a domed town/city. ACE-Certified applies to the entire lifecycle of each product and/or service (food, shelter, clothing, etc.), including how we produce, distribute, operate, repair and recycle them. ACE-Certified products and services are designed to operate both inside and outside of the town/city domes.
    2. Once Stage 1’s infrastructure is ready and proven safe for use within theoretical city/town domes, build the domes. Start off with shopping mall domes, ensure the ACE-Certified infrastructure enables the shopping mall to be a completely closed ecosystem and pollution-free. Then expand to partial town domes, then full town domes, then partial city domes and finally big city domes. Once we have a few city domes operating problem-free for a decade, then we can progress to Stage 3.
    3. Underwater habitats and L5 space habitats present entirely new challenges and dangers that must be overcome, but in which escape to safety is still quicker and easier than if we are travelling through or living in interstellar space, if a failure in the system occurs.

    Once these stages have been mastered, the transition to space exploration and permanent space colonization elsewhere in our solar system and beyond will be infinitely easier and safer than without them. My website provides greater detail.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>