In my story, a ship is built in orbit around Earth, designed for a 20-person scientific expedition to a planet 11.4 ly away. It will accelerate at 1g for half the distance, then flip around and accelerate at 1g in the opposite direction until it comes to a stop. It's not designed to ever enter an atmosphere.

Technology in this world is arbitrarily powerful within known laws of physics, and they're able to produce large quantities of antimatter and perfectly convert that fuel's rest mass into kinetic energy via a gamma ray drive. In fact, they can even make solid antimatter, so if the fuel source was made of something like Bismuth and anti-Bismuth, it could take up very little space. They can also harvest matter for fuel at their destination.

I assumed that the habitat they'd need for the journey would weigh 0.5 million kg (this is comparable to the mass of the ISS, and while obviously this would need to be much larger and self-sustaining, it's also made with futuristic ultra-light materials) This handy calculator tells me that the total mass including fuel will come out to around 100 million kg.

This sounds like an absurd number until you realize that the Empire State Building weighs about 3 times that much. Piece of cake for an advanced civilization. (Did I mention they can get the materials to orbit with a space elevator?) More specifically, the mass is equivalent to 10,000 m^3 of Bismuth, or a 10x10x100 m building's worth, plus the habitat on top (some of the material can double as a radiation shield during the journey).

Given all of that, my question is simple: What would the ship look like?

My thoughts so far (feel free to contradict me if I made an incorrect assumption):

  1. The ship looks more or less like a skyscraper. After all, it's roughly skyscraper-sized, and it needs to hold itself up under 1g of acceleration, just like skyscrapers on Earth do. But the lower 100 floors or whatever will be fuel.
  2. The bottom of the ship is a lattice of parabolic dishes. The idea is that the gamma rays will be produced at the foci and reflected to generate thrust. However, I'm not convinced that a material capable of reflecting gamma rays is even physically possible. If there's an alternative, what is it and what would it look like?
  • 2
    $\begingroup$ If the engines turned off when the ship hit 90% lightspeed, you'd use dramatically less fuel and still arrive at almost the same time. However the crew would have to live without that sweet 1g environment for a while. $\endgroup$
    – Robyn
    Nov 1, 2019 at 21:18
  • 1
    $\begingroup$ Okay, just did the math on that. Coasting after 0.9c makes the entire trip take 13.9 years instead of 13.2 (so, just eight months longer), but from the perspective of the crew, the length goes from 5.1 years to 7.2, and 4.3 of those years (most of the trip) will be in 0g. That's a rough tradeoff. Though admittedly there is a 90% reduction in fuel, so your proposal is probably the way to go. I wanted to avoid a centrifugal gravity ring on the ship, but with those kind of savings and that long in 0g, it might be necessary after all. $\endgroup$ Nov 1, 2019 at 21:49
  • $\begingroup$ Why flip that ship if you just want to reverse thrust? $\endgroup$
    – syck
    Nov 1, 2019 at 22:29
  • $\begingroup$ @syck Am I missing something? The thrust in spaceships comes out the bottom. You have to turn the ship around to get it to point the other way. Plus, this way, the direction of apparent gravity always points downward through the floor of the ship, so you don't need to make sure that the ceilings work equally well as floors and move furniture around and such. $\endgroup$ Nov 1, 2019 at 22:52
  • 2
    $\begingroup$ I just don't see why it's worth the trouble. Rotating a spaceship while it's moving at constant velocity is a really simple thing to do, even if it has a lot of mass. You need to have systems that can rotate the ship anyway for steering reasons. $\endgroup$ Nov 1, 2019 at 23:46

3 Answers 3


(editted for brevity, if you can believe that. the edit history retains some worked analysis of the rocket performance and a comparison to the Frisbee antimatter starship design)

TL;DR: It will look a little more like this than like a skyscraper:

ISV Venture Star

(Note the presence of heat rejection, the glowing red bits around the rockets, debris shielding, the shiny rhombus at the other end, and if you squint a bit, the spun gravity sections near the shielding. All hallmarks of hard-scifi design. More details in this diagram)

This is the ISV Venture Star from the film Avatar. It is based on an older work called Project Valkyrie which involves turning the traditional rocket design upside-down, using a pair of carefully angled rocket engines at the front of the ship, and a fairly long tension structure which supports the payload of the spacecraft. The tension structure is simpler and lighter than a compression structure, and this means you can make it long enough so that sheer distance from the intense radiation sources of your rocket engines helps protect you (thanks to the inverse square law) and reduces the mass of shielding you need to keep everyone alive. The Venture Star was only 1.6km long, but the Valkyrie design was more like 10km.

The Venture Star also uses a laser sail to boost it away from Earth, and to slow it back down again when it returns, which reduces the delta-V (and hence fuel mass) needed for each leg of the journey by a factor of 4. You might not want to use a laser sail for whatever reason, but you might consider using a magnetic sail to slow you down for the deceleration phase of your journey. This lets you slow your ship by magnetic drag against charged particles in interstellar space and the solar wind emitted by your target star. This sort of parachute can brake with considerable force (the example given in the link starts out at over 5 gravities) and could brake a ship from .95c to .01c in a couple of years with no expenditure of propellant.

A combination of these approaches might allow you to either a) not use any rockets at all (requiring you to build a laser array at your destination) or b) to fuel up your rocket at Sol, then be boosted away by laser, brake by magnetic parachute, and then use your rockets to boost you up for the return leg of the journey meaning that you wouldn't need to build an antimatter fabrication facility at your destination (and this may also avoid the need for magical handwavium ultra-efficient antimatter synthesis, fractionally increasing the plausibility of your work).

Now for the more detailed grumbles:

In any case, there are some problems that your initial analysis missed:

  1. Mass. You have a ship that weighs 500 tonnes at its destination, but 100000 tonnes on launch, with all of that extra mass as fuel. This implies your fuel tanks can confine millions of kilos of antimatter, but weigh only grammes, or that you can perfectly cannibalise each fuel tank as the matter-side of the antimatter rocket reaction. It also implies that an engine that generates hundreds of petawatts of power also weighs an insignificant amount, which is extremely implausible.

    Using lightsails to start and magsails to stop helps with the mass ratio issue. It also reduces the inconvenience of bulk antimatter production, and the safety issues associated with storing vast quantities of the stuff. If you really must use self-powered flight, then coasting for much of your journey isn't going to take subjectively much longer and building centrifugal gravity facilities will be easier than making and storing millions of kilos of antimatter for years at a time.

  2. Engine radiation shielding. Antimatter rockets emit vast amounts of gamma rays (your ship might have an initial flux of some 200 petawtts of the damn things), many of which will impact your ship. If they hit your antimatter, they risk causing a chain reaction that will annihilate your ship. If they hit your crew, they'll die. Even magical shield generators weigh something... the shielding for your ship, your rocket engine and your fuel isn't going to be massless. Even the long-tether Valkyrie/Venture Star design won't protect the rockets themselves, and you still need to get fuel up to them so the fuel tanks and lines will still need shielding too!
  3. Debris shielding. At 99% of the speed of light, a single gram of dust at this speed packs about 123 kilotonnes tnt equivalent. Even in the pretty vacuous interstellar medium around our sun, you'll find neutral hydrogen and helium atoms, tens of thousands of them per cubic metre, and at top speed each square metres of your ship's cross section is sweeping out over 296 million cubic metres. This paper suggests that an unshielded human travelling over .6c might expect to receive about 104 REM per second, which is over 50 times a lethal dose of radiation. Every second. I'm pretty certain your 500 tonne ship does not carry enough shielding to protect against this sort of punishment... you'll need several tonnes of water per square metre of shield. You might be able to use some of your rocket fuel for this, if you were careful.

    Debris shielding can be partially handled by clever solutions to heat rejection... the Valkyrie design uses a "fountain" style liquid droplet radiator, spraying hot coolant into space ahead of your ship, and letting the ship's own acceleration catch up with the drops to recover them. Some of the drops will be scattered by incoming gas and dust particles so you'll need to bring spare coolant, of course. There are many other approaches, but their details are out of the scope of this question.

  4. Heat radiation. Those gamma rays need reflecting or absorbing. Unless you have the most magical of shields, the reflection process is going to need power and generate heat. The drive coils will need powering and cooling. The cryogenic refridgerators for your antimatter storage systems will need powering and also need huge heat sinks. The list goes on... you're dealing with a multi-hundred-petawatt reaction engine, and even with phenomenal efficiencies a suitable liquid droplet radiator array might still be several square kilometres in size, and will necessarily weigh thousands of tonnes that you haven't taken account of. The mass of those heatsinks will need more rocket power, fuel and reaction mass to push them, and more shielding to help keep them intact.
  5. Engine design. Antimatter rockets are even less efficient than you might think, because some of your propellant is evaporating promptly, and the mass-energy of the rest is lost into space. Even with magical gamma ray reflectors, your mass ratio will be higher than your initial analysis implies. Robert Frisbee wrote an interesting paper on antimatter-driven starships (worth a read; it uses no sail technology and a skycraper design, and as a result is really, really huge, like hundreds of km long and a lot slower and simpler than your design) which suggests it could need 4 to 5 times the mass ratio than a classical rocket design of similar performance would need. This drastically limits your practical delta-V, making boosting up and down to high relativistic speeds even more vastly unlikely.

    If you can handwave the production, storage and pumping of neutral, ferrous antimatter then maybe you'll be able to get away with a beam-core antimatter rocket, but practically the cross section of cross sections of particle/antiparticle collisions is very low. Your rocket's reaction chamber will have to be very long to ensure it all gets burn up, and that length makes forming a rocket nozzle or reflective photon rocket system very difficult indeed, and increases the issues associated with shielding and cooling.

    The author of the that paper (and indeed the radiation effects on relativistic starships paper) suggests using an antimatter power reactor and ion drives instead. The tech level goes down a little, and the plausibility goes up. Of course, if you are using light and magnetic sails, then you can just use vastly simpler antimatter-catalysed fusion rockets instead for terminal phase braking and in-system manoeuvering.

  • $\begingroup$ Okay, some really good stuff here. I had no idea Avatar had such a realistic starship, and the tether idea seems really cool. Will probably use. For gamma rays, I actually have some handwavium in the story that allows me to convert most of them into charged particles (electrons and positrons), which should drastically alleviate both the waste heat problem and the radiation risk, and give me more thrust to boot. I also hadn't considered the possibility of detaching and rearranging parts of the ship. That should allow me to keep the dust shield in front while the ship reorients for deceleration. $\endgroup$ Nov 2, 2019 at 15:54
  • $\begingroup$ While laser propulsion is awesome, something else that I didn't bring up is that the planet they're visiting is uninhabited, and they're only staying for 5 years, and they're trying to cause minimal disruption to the planet's environment. That makes it unlikely that they'll be able to deploy a laser for the return journey unless they bring it with them (and I suspect the math doesn't work out for that being a good use of mass). I think with the aforementioned handwavium, I can make antimatter rockets work out (thumbs up for magnetic drag sails, though. Good way to cut down on fuel). $\endgroup$ Nov 2, 2019 at 16:12
  • $\begingroup$ @GiladM how are you intending to make the best part of a hundred million tonnes of antimatter at an uninhabited world in a few short years? The infrastructure required is unlikely to be vastly simpler than a big laser array. $\endgroup$ Nov 2, 2019 at 17:28
  • $\begingroup$ @GiladM and whilst I think of it, a rigid, vanishingly low-mass perfect broad-spectrum gamma ray mirror implies either a) energy shields or b) an entirely new form of material that isn't made of regular particles. That in turn has a lot of serious knock-on effects, regarding science, technology and warfare. It also makes it even more likely you'd use laser or other beam driven propulsion, if only for the boost phase of the outbound leg of the journey. You could still carry enough antmatter for the boost phase of the return journey, of course. $\endgroup$ Nov 2, 2019 at 19:17
  • $\begingroup$ I agree, the gamma mirrors won't be there. I was looking for a way to avoid them, and I found one: turn the gamma rays into charged particles and repel them. As for making fuel on the other world, I'm still fuzzy on the details, but I think I'll go with Robyn's suggestion and turn off thrusters once at 0.9c to save fuel. That cuts the mass by a factor of 10. In this world, antimatter is ridiculously easy to make (the handwavium I discussed earlier), but I agree that 100 million kg would be pushing it. $\endgroup$ Nov 2, 2019 at 20:27

Your first point is great. Spacecraft that are laid out as if there's a "down" that's perpendicular to the direction of thrust are almost always a stupid design and a holdover from depictions of spacecraft as essentially just aircraft or naval ships. So yes, laying the ship out as a stack of consecutive floors, sort of like a skyscraper, is objectively the best design for this kind of spacecraft. However, you've got a couple of issues here.

1. You're going to need a way to store that antimatter.

Your antimatter is, for some reason, in the form of anti-bismuth. OK, the first thing we're going to need to do is to give it an electric charge. We can't just hold it in a tank made of matter, because as soon as it touches the walls, your solar system gets a temporary second sun. So we need some way to hold it in our ship without ever touching it. Luckily, there is a way to do that (although it only works with diamagnetic anti-elements, bismuth is luckily the most diamagnetic element, of all, so good choice of fuel). Your incredibly advanced civilization undoubtedly also has incredibly advanced magnetic fields, so if we give our anti-bismuth a charge, either by adding or removing a whole bunch of positrons, you can contain it in a "tank" made of powerful magnetic fields. You'll then need to surround the whole thing with an actual, physical tank to prevent the interstellar medium, which will be hitting you at >0.9c, from slowly eroding your antimatter, but as a super-advanced civilization, I doubt building a big, absurdly durable fuel tank with a bunch of carefully placed and tuned electromagnets inside will be a problem. Additionally, since your magnetic fields are going to need to hold a huge mass of anti-bismuth against 1g of acceleration, they'll need a ton of energy input, but since your ship runs off antimatter, I doubt electricity generation is much of a problem. Note that this also means that you'll have to allocate more space for your antimatter storage than your normal fuel storage, so make note of that when drawing up designs, although how much extra room you need depends on how advanced your civilization is. You're also going to want to store both your bismuth and your anti-bismuth as liquids, to allow you to actually use them as fuel without having to somehow chip off chunks of solid antimatter. This will require keeping them nice and warm, above their melting point, so don't forget to include some heaters in your design.

2. You're not just going to produce gamma rays (but that's a good thing)

Contrary to popular belief, if you smack a chunk of matter into a chunk of antimatter, you don't just get energy. (well technically, if you smack a positron into an electron, all you get is gamma radiation, but since we're talking about anti-bismuth and bismuth here, we're going to have anti-protons and anti-neutrons involved too). You'll get some gamma radiation and a bunch of high-energy particles, some charged, and some uncharged. Unfortunately, you're probably going to have to waste the gamma rays, as we don't actually have an effective way to utilize them because as far as we know, nothing reflects gamma rays. The thing you're going to be using as thrust will have to be the charged particles (pions to be precise, not that it really matters), which will make up about two-thirds of the energy from the annihilation event, with the other one-third being those wasted gamma rays. You'll have to use a magnetic nozzle to direct them, which will probably look like a wide, thin, curved ring attached to the back of your ship with a few support struts. So that's what the bottom of your ship is going to look like, rather than a bunch of parabolic dishes.

If you handwave in a material your society has than reflects gamma radiation, feel free to build a physical rocket nozzle out of that to surround the magnetic nozzle (this would probably just appear to be a standard rocket nozzle, rather than some sort of lattice of parabolic dishes), but be aware that that's stepping pretty far outside our current understanding of physics. Everything I've laid out in my answer has been "super-advanced, but definitely possible", a gamma-ray mirror is definitely outside that realm.

So, assuming your advanced society lacks any handwavium gamma-mirror, your ship will look like kind of like a skyscraper, especially near the top. The very top part, where the crew lives, will probably be rectangular or square in shape, to give the pesky humans their nice, regular rooms, and, depending on the height of your spacecraft, may include some radiation shielding seperating it from the rest of the ship to prevent gamma rays from the drive giving everyone cancer, although the intervening mass of fuel helps with that, and as it is used up, the amount of radiation decreases as the drive power is turned down to maintain constant acceleration, so that works out nicely. The fuel tanks, below this, will probably be cylindrical, to save on tank mass and the number of electromagnets required to keep the anti-bismuth contained. These tanks will likely be stacked with the bismuth tank above the anti-bismuth tank, to simplify the already horrific engineering involved in transporting the anti-bismuth to the engine without letting it touch anything. The engine, at the bottom, will appear to be a wide, thin angled ring (or possibly several concentric ones), attached to the hull by struts. Inside this invisible magnetic nozzle, bismuth, carried from the upper tank, will annihilate with anti-bismuth from the lower tank and create a flood of gamma rays and charged pions. The pions will be funneled away from the spacecraft by the magnetic nozzle, providing thrust, and the gamma rays will spread out randomly in all directions, and hopefully not give any of your scientists cancer. There may be some visible light from the interaction site, and surfaces at the base of the ship may glow as they give off the energy they absorb from gamma radiation, but there will be no visible drive plume.

  • $\begingroup$ Great points! (I kind of feel like a moron for not mentioning that the antimatter has to be isolated from the matter. I swear I didn't just forget). I like the charged pion idea, though apparently that lowers my energy efficiency from 1.0 to 0.664, since that's how much of the reaction energy is in the form of charged pions. Oh, and I just said Bismuth because it's the densest non-toxic, non-radioactive material. It doesn't have to be Bismuth. They have nuclear alchemy; they can make arbitrary quantities of any matter to use as fuel. I'd be interested to know if you think of something better. $\endgroup$ Nov 2, 2019 at 9:35
  • $\begingroup$ Re: gamma rays, while we can't reflect them, we can absorb them, which, iiuc, transfers their momentum to the absorbing material? So it wouldn't be as efficient as visible lite or microwaves, in terms of momentum transfer, but some of that should still contribute. Even if it's all converted to heat, the heat should be re-emitted as lower energy photons? $\endgroup$
    – CAE Jones
    Nov 2, 2019 at 13:54
  • $\begingroup$ The pions contain the remaining 2/3rds of the energy of the annihilation reaction in a combination of their kinetic energy and their mass energy. You can turn the kinetic energy into thrust via a magnetic nozzle, but what's left in their mass energy is effectively wasted, as far as rocketry is concerned. This is all part of why the effective exhaust velocity of a beam-core rocket is more like .33c. $\endgroup$ Nov 2, 2019 at 21:14
  • $\begingroup$ @CAEJones there are a few problems there... one is that gamma ray absorption can result in ionisation or photodisintegration, and those lost bits of atom ain't coming back if they're part of a rocket engine. Another is that you need a pretty vast emitter area to release all that heat in the right direction without having anything melt (IR photons carry vastly less energy than gamma ray photons). $\endgroup$ Nov 2, 2019 at 21:24

Yes - it would be a skyscraper, and you need a way to withstand those pesky interstellar dust particles, and radiate away heat

Regardless of your drive and your power source, constant 1g acceleration would slowly increase the speed of your ship such that even after just a half a lightyear any dust particles would hit your ship with enormous force, and let alone for a 5.52 ly journey (with the same problem on the deceleration part).

Not to mention the constant ablative impact of hydrogen particles, or anything more substantial.

So you would need a heavy 'top' shielding to your skyscraper. And your skyscraper should be as 'thin' as possible to reduce the cross-sectional area exposed to the dust and interstellar medium.

Impacts would also create a lot of heat, and in space heat is hard to get rid of.

Not only that, but the critical-danger part would be the 'flipping' of your craft at the half-way point. Add to this your engine would need to be facing the impacts to decelerate the craft on the deceleration leg, and would be exposed to the dust too.

Possible solutions:

  • Have sacrificial bodies ahead of your habitable spacecraft, such that they would absorb the impact of interstellar dust prior to any craft that may endanger life.

  • Reduce the cross-sectional area to just the size of a small house, perhaps 150sqm maximum. You can still have lifts to allow people to interact, for the journey would still be several years. You can have simply 2 lift shafts, one up one down, with the ability to 'park' lifts to allow them to bypass each other. This would mean your ship would be very very long (or high...) but would still be functional.

  • Detach the shield on the deceleration leg, and 'push' against it with your drive, to enable the drive to operate without exposure to the dust, and perhaps with greater efficiency on the deceleration leg.

  • Find a way to radiate heat forward, to assist with getting rid of objects in your path too. Perhaps a heat exchanger heating very hot 'pellets' to fire in front of you, or beam it somehow forward.

  • Modularise each level of your ship, such that at the critical 'flip' point you are not trying to rotate in a pinwheel way (which will likely catastrophically rip your ship apart), but to rotate 'modules' inside the body of the ship instead to retain it's needle-like aspect ratio.

It's an exciting prospect to design a ship like this - I have always though this is the best solution for interstellar travel, obtaining the speed required to reduce the time needed, and get the gravity needed, in one solution. Keep letting us know how you go with your design work...


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .