Spaceships of the Expanse

I have been enjoying “The Expanse” series by James Corey. It’s a space opera set a couple hundred years from now, after humans have colonized and populated the moon, Mars, the asteroid belt, and outer planet moons. Spaceships journey between these worlds, complex engineering projects remake asteroids into habitable stations, and space navies boost from place to place to fight space pirates. I think it’s great because it captures what I wish for humanity’s future: that we will go out and colonize other worlds, that we will be able to undertake engineering projects for the greater good, and that we will become robust enough to weather grand challenges – things we see in the world today as global warming, income inequality, nuclear proliferation, and the like. In many ways, the first three books are about the tension between such grand visions and idealism, and politics and profiteering.

Leviathan Wakes cover, from Orbit

The books are soon going to be a TV series, and I am very much looking forward to see its depiction of space and space travel. (With the exception of parts of the first book, wherein Corey tried to write something horror-ish by being as gross as he could think to be. Whatever. Those are not the good parts of the book.) Corey steered clear of many sci-fi tropes that would have a big impact on the appearance of the series – no artificial gravity here! – and he made sure to build aspects of spacecraft engineering and operations into the cultures he depicted. For example, “Belters” nod and shrug with whole-arm gestures, so that they can be seen when wearing a suit. A good chunk of the books take place in zero gravity. Hopefully that will translate to the screen!

I’m going to take a look at some of the spacecraft engineering concepts in “The Expanse.” Let’s start with the most science-fictional, and therefore least plausible:

The Epstein Drive

Corey very quickly establishes that the powerhouse of his whole solar-system-wide civilization is the “Epstein Drive,” which is some kind of fusion engine for boosting spaceships around. It allows craft to thrust continuously from one planet or asteroid to another, accelerating constantly for half the trip and then decelerating constantly for the second half. This trajectory allows relatively quick travel times between worlds. Conveniently for crew health, and for TV production, the engine also provides “thrust gravity” inside the spaceship. Ships are therefore designed with decks in “stacks” above the engine with a ladder or lift giving crew access between decks, like in a skyscraper.

A fusion engine isn’t a crazy idea, especially not for a civilization a couple hundred years in the future. The problem is propellant. No matter how powerful or efficient your engine is, you will always need to be chucking propellant out the back to sustain this kind of thrust profile.

Picture this: you’re sitting in the middle of a frozen pond. The ice is perfectly frictionless, so you can’t walk or crawl or anything to get back to solid ground. What you do have is a bag full of baseballs. If you throw a baseball away from you, then you have given it some amount of momentum (mass times velocity). Your body gets an equal and opposite amount of momentum: you start sliding in the direction opposite your throw, but much more slowly than the baseball (because its mass is small while yours is big). Great! You have a way to get to shore. But you don’t want to wait out this long slide, so you throw another baseball. This speeds you up a little. Another throw speeds you up a little more. You can keep throwing baseballs until you decide that you’re going fast enough that you can wait it out. That’s basically how spacecraft work now: they thrust for a little, and then coast for a long period of time until they get to their next destination. But what if you wanted to keep thrusting the whole time? You will need more baseballs. Lots more baseballs. You are going to have to keep throwing them, constantly, to keep accelerating yourself.

Writing that a spaceship has a fusion drive instead of a chemical rocket is like replacing yourself in this analogy with a major league baseball pitcher. They will put more momentum into each pitch, and so they’ll go faster across the ice. In other words, their thrust is more efficient. But they will still run out of baseballs at some point, and then they must coast without thrust. The spaceship must stop its burn, cease thrust gravity, and wait several more months before getting to their destination. In the end, high thrust – and, with it, appreciable thrust gravity – should only be active for a short time in any space voyage through the Expanse. As we are learning with ion propulsion nowadays, it can often be most efficient to run at a low level of thrust, but sustain that for a very long time. But that doesn’t give our characters a convenient floor to stand on! So Corey put the word “Epstein” in front of “fusion drive.” “Epstein,” in this case, is short for “magic.” It’s a kind of magic that lets Corey have thrust without propellant, so that he can simultaneously achieve short (astronomically speaking) travel times and keep his crew in thrust gravity.

For a more physically realistic depiction of relationship between fuel, propellant, and thrust, consider Neal Stevenson’s spaceship Ymir in Seveneves.

The Way Ships Move

In the Expanse universe, spaceships are flipping around all the time to vector their engines in the correct direction to change their velocity. And we often read references to what the thrusters are doing on ships. This is all good. But the ships don’t really move the way real spacecraft move.

First of all, orbits barely enter the picture. One scene in Leviathan Wakes involves a character plotting out the likely trajectories of a certain ship, but other than that, the characters can go just about anywhere they want to go as long as they have a good ship to call theirs. Absent the Epstein “magic,” that behavior isn’t really plausible.

Second, though, is that Corey imagines his spaceships rotate themselves around in the same way just about all science fiction authors do: with thrusters. That’s not what most modern spacecraft do! They actually use wheels. Spin a wheel clockwise, and Newton’s third law kicks in: there’s an equal and opposite reaction. The spacecraft spins counterclockwise. Devices that function as I just described are, therefore, called reaction wheels. Other wheel-based devices that take advantage of gyroscopic torques can give satellites quite a lot of agility – without using any propellant. I suspect that the reason why these realistic actuators don’t often appear in science fiction is that there are no obvious cues to their operation: no thruster spurts, no blue glows shining out of emitters, nothing. The ship just starts to rotate.

I was happy to read that Corey’s spaceships are all native to space. There are not many cases where a ship lands, and in those cases, it’s always a small one. The heroes’ ship does once, on Ganymede. With surface gravity comparable to Earth’s moon, that’s not such a stretch for a fusion-drive starship.

28 June 15 Edit: Darn it, I just started Cibola Burn and about the fourth thing that happens is that the Rocinante lands on an Earth-size planet and immediately takes off again. Minus points for that!

The Battles

Space combat plays a big role in the plot of the Expanse books. And it’s a generally great depiction of space combat! Lots of the tactics and technologies are grounded in plausible physics. Ships shoot missiles and guns at each other, the effective range of a torpedo is determined by how close your ship needs to be to make sure the enemy ship doesn’t have time to shoot your torpedo down, the crew all gets into space suits at the beginning of the battle, there’s a ton of electronic warfare activity, and the battles wax and wane in intensity as the spaceships maneuver and orbit.

I’ve long thought that the most effective weapons in a space battle would be simple kinetic slugs or flak shells. My reasoning is simple: the speeds of objects in space are fast enough that a relatively small piece of junk can easily blast a hole through sensitive components. This is exactly why present-day spacecraft engineers – like me – worry about micrometeoroid strikes, space debris, and the Kessler syndrome. In the Expanse, the ships all fire torpedoes or guns at each other. And the results of weapon strikes are devastating: it only takes one torpedo or a few well-placed railgun slugs to take out a ship. Ships blast electronic garble at each other to screw up their targeting systems, but in the end the best defense is not getting hit – so we see the pilot do a lot of evasive maneuvering. I think this is all on the right track from a physics standpoint, though a real space battle with “Expanse-style” ships would probably take a lot longer, involve more orbit dynamics, and require a lot more computerized coordination.

There are two rather implausible elements to the battles. First is the Epstein Drive, which makes the combatants’ maneuvering matter a lot more than orbit dynamics. Second is the “juice,” a drug cocktail that keeps people alive and functioning when exposed to high gee forces. As a way to deal with high gees, the “juice” is just about as good a science-fictiony way to do it as any other, including immersing people in fluid as in The Forever War or inventing some kind of mythological inertial dampener. In the end, though, humans are squishy, precious cargo, and fighting full-on battles with them inside your spaceships doesn’t make a whole lot of sense.

Stealthy Spaceships

(There are some minor spoilers in this section!)

A plot point early on in the first book, Leviathan Wakes, revolves around the appearance of a stealth spaceship. This doesn’t involve any cloaking devices like in the Star Trek universe. Rather, a few spaceships avoid detection by (1) being painted black, which hides them in the visible spectrum, (2) having surfaces that absorb or scatter radar, which hides them in radio wavelengths, and (3) radiating heat out the side of the ship facing away from the enemy, which hides them in infrared. Much as it might give some people heartburn, this is all fairly plausible! The first two points are easy to imagine based on what we know about about the present-day Air Force. Though its not as familiar to the general public, the third item is actually something that comes up all the time in spacecraft design: especially if your satellite has sensitive electronics – like an infrared telescope – the design will include coolers, heaters, baffles, insulation, and radiators designed to emit heat in directions pointing both away from the precious detectors and away from the sun. Even the International Space Station has radiators that rotate to keep them pointing away from the sun most of the time. (The reason is that if the radiators face the sun, they’ll start to absorb heat into the station instead of emitting it out!) Such a “thermal management system” could be designed to, with the other stealth elements, give one side of a spacecraft the appearance of a cool, black spot indistinguishable from the rest of empty space.

A stealth spaceship wouldn’t be easy to build, and it wouldn’t be perfectly invisible – just harder to detect than normal to a lone adversary. And, in fact, both those points are relevant to the spaceships in the Expanse. One crewmember is able to spot a stealth ship on his sensors, but he doesn’t know what it is or how to respond to it. And that’s really all it takes for the stealth ship to accomplish its mission, after all! The very difficulty of constructing a stealth spacecraft actually makes the stealthiness more effective. The characters who cannot conceive that somebody could field a stealth spaceship end up more prone to falling prey to it.

Spaceship stealth makes an appearance other times, as well. At least twice in the series, the heroes’ ship hides itself by masquerading as something it’s not.

Spin Gravity

Lots of space stations in the Expanse, including some embedded in asteroids, spin to provide their inhabitants with centrifugal “gravity.” This is an idea that’s been around the aerospace and science fiction communities for decades, and Corey executes it well. In fact, one of the things I enjoy about the books is how the plot moves between the different environments of planetary gravity, low lunar gravity, spin gravity, and (“Epstein”-based though it may be) thrust gravity. The different gravitational environments contribute to different cultures, and they put the characters in interesting and different situations. If the TV series sticks to the books, we’re going to see low-gee gunfights and damage control teams solving problems in microgravity. Regularly.

All the stations with spin gravity are large, which is the right choice. It means they don’t have to spin at a dizzying rate to get a comfortable level of gs for their inhabitants. There is another benefit, in that the weird non-intuitive kinematics of rotations – Coriolis forces – have less of an effect the larger the rotating space station. These effects can be truly weird, and it can be hard even for physicists to bookkeep all the terms correctly to model them. Something that might happen if you were standing in a spinning space station is that if you drop something, you will actually see it follow a curving path to the floor, and it won’t land at your feet. You will also see it fall at a different speed than you would expect based on the gravity you experience. (I’m planning to write something up separately to go into all the details.)

Anyway, suffice it to say that spin gravity is a strange environment and Corey, like most science fiction writers, doesn’t go into all the details. But spin is the right idea for giving gravity to spacefarers, and I can’t wait to see how the visual effects team on the Expanse interprets all the spinning structures.

All in all, I’m thinking that The Expanse will be good for science fiction on TV. It will be a show with a time period a bit closer to us than, say, Star Trek. And the show will have a wide diversity of environments to challenge the characters. I am looking forward to seeing their depictions of spacecraft and how they move around in space!

35 thoughts on “Spaceships of the Expanse”

  1. Good article, though I have a couple of comments.
    Modern spacecraft (Dragon, Orion) do use reaction control thrusters in order to rotate. Reaction wheels are used on ISS, but it still needs thrusters to dump the angular momentum every now and then, because the wheel has its limits.

    If you want to turn a huge spaceship fast (combat situation for example), thrusters are probably the best choice. Otherwise you’d need a very heavy reaction wheel.

    Orbits aren’t much mentioned because the spaceships are almost always flying so fast that they’re in Solar escape trajectory, hardly curved at all (thanks to Epstein drives). Of course, when “parking” near a station or a planet, they have to slow down and orbit, which is mentioned in the books I think.

    1. Karrizfin – Dragon and Orion use thrusters, yes. But the hundreds of other spacecraft in orbit use reaction wheels or control moment gyros (like ISS, which has four dinner-table-sized CMGs). Agile imaging satellites use CMGs, which can produce a LOT of control authority without the disadvantages of thrusters. You’re right that there have to be periodic “momentum dumps.”

      1. I always imagined it was a combination of both in Corey’s universe. CMGs work extremely well, as mentioned, but have their limitations (also mentioned). In terms of being able to make quick turns in any direction at any time, it would make sense to employ a combination of reaction control thrusters and gyros. This is also provides some redundancy in the event of damage (plus having a CMG hit its “wall” in the middle of a combat maneuver could prove fatal – re: momentum dumps etc.).

        Also take into account the mass of combat ships in relation to the agile imaging satellites you mention. These satellites are designed for minimum mass. I would imagine one torpedo on a ship like the Rocinante would weigh as much if not more than an entire scientific satellite, when you combine the mass of the warhead(s), fuel, engine, etc. on the torpedo. Granted there doesn’t seem to be any mention of CMGs or reaction wheels in the books, and Corey is usually pretty good about the details, but I would imagine that spacecraft as complex as these would implement multiple attitude control systems.

        1. As far as I understand, CMGs and reaction wheels are used for specific purposes, such as fine control in satellites or ISS attitude.

          They can also be unloaded via Earths magnetic field or the gravity gradient, so no fuel is needed.

          Even if they have to be unloaded via thruster, I imagine for reliability etc it’s better to do fewer but longer thruster burns to unload them, vs lots of tiny burns for attitude control.

          IMO for spacecraft that don’t need ongoing or fine control, or don’t have a thruster free way to unload momentum, then a CMG or reaction wheel is not very useful.

          In the Expanse universe, I can’t see CMGs being worth the extra fuel cost to haul around except in limited specific circumstances (backup combat systems is one). Even if they can be unloaded efficiently, the ships just don’t need to change attitude enough for it to be worth it.

          Does anyone know enough about CMGs to calculate how much they would weigh, in a configuration that could change attitude on the Roci fast enough to be useful as a backup system for combat?

          1. Good points. Would be interested to see those calculations as well. I don’t even know where one would find specs of the Roci to begin such a calculation.

          2. Actually, I know enough about CMGs to make the calculation! I will try and find some reasonable specs or assumptions for the Roci – dimensions and inertia, or if not inertia at least mass.

            One quick calculation: The International Space Station’s CMGs can produce a torque of 260 N m (source: this paper). This page refers to a cold gas thruster (a thruster that “runs teakettle”) which produces 111 N of force at a specific impulse of 68 s. Placing that thruster with a moment arm 2.34 m from the Roci’s center of mass will produce the same torque as the CMG. However, the thruster will consume propellant at a mass rate of (111 N)/(g*68 s) = 0.167 kg/s. You can increase the torque from the thruster with no mass consumption penalty by moving it to a larger moment arm. You can similarly increase the torque of the CMG by making it spin faster, or giving it a bigger inertia. I will try and come back with something on the mass of the CMG system equivalent to the best thruster configuration I can come up with.

            The only other point I would make to the comment above is that CMGs and reaction wheels are not the actuators used “for specific purposes.” They are THE baseline actuators on modern spacecraft, precisely because they expend no consumables and have a sufficiently large torque capacity. On vehicles designed for long-term operation in space (i.e. not ISS cargo shuttles) thrusters are the backup actuator, used only to desaturate momentum built up in the wheels or in the event that something goes wrong. Many spacecraft don’t even use thrusters for momentum dumping, they do it with magnetic coils, making thrusters well and truly the backup actuator.

            Then again – all that is true of CURRENT spacecraft! We can’t fill up their propellant tanks with comet chunks. 🙂

          3. To be fair, running teakettle is using a fusion reactor to (presumably) superheat water to steam, which is a totally different league to compressed gas thrusters. Likewise, future CMGs would have higher limits than current ones.

            One limiting factor (especially for a military ship) could be the peak gforces on the crew, or ship structure as they spin.

            “CMGs and reaction wheels are not the actuators used “for specific purposes.” They are THE baseline actuators on modern spacecraft, precisely because they expend no consumables and have a sufficiently large torque capacity”

            CMGs and reaction wheels are not actually a baseline actuator though, precisely because they are limited in use to specific purposes and circumstances.

            Expending no consumables is probably the key example. Importantly, it’s actually expending no consumables in certain situations. I am sure CMGs and reaction wheels would be a baseline actuator if they could provide attitude control without expending consumables in any (or even most) situations.

            CMGs are common yes, but not universally useful. Going beyond orbital space stations and satellites (where dumping momentum is easy), CMGs appear to have comparatively fewer advantages.

            Some of our longest operating, furthest ranging and most modern spacecraft don’t use CMGs, because the disadvantages outweigh the advantages. In contrast, CMGs offer many other spacecraft in different situations very important advantages.

            For any spacecraft, (theoretical or existing), the advantages and disadvantages of CMGs need to be weighed up against the expected mission – they are not useful by default.

            Which brings us back to the question at hand, would the Roci use CMGs? As far as I can tell, the key advantage of consumable free attitude changes don’t apply. Being a warship, battle redundancy could be important, if CMGs offer the performance required (which I don’t know).

            @josephshoer What would you say are the advantages of CMGs to Roci?

          4. When reading about ISS zero propellant maneuver, I came across some interesting numbers.

            50.8 KGs of propellant was used to rotate the ISS 180 degrees, which the CMGs can do without consumables.


            Which is awesome. But what if we strap a Epstein drive to the ISS (heh) and start flying missions around the solar system?

            The CMGs weigh 1088 KG all up. We need thrusters for translation control anyway, so there is no weight penalty to having them.

            So in the most basic sense, our ISS ship can do at least twenty 180 degree rotations, for the same weight as bringing along the CMGs.

            (Ignoring that there is less weight to carry each time the thrusters are used) (Also some sources suggest that the thrusters do it a lot faster, but I can’t find the specifics so don’t include that)

            This is super simplistic, but I think a reasonable / interesting starting point to considering if we want to add CMGs to our ship. To keep it simple, I am assuming that the CMGs never need to unload momentum and that all other costs are inconsequential compared to fusion fuel. Specific economics of the situation could change things a lot.

            For a lot of the journeys we see in The Expanse, the ships don’t need to make many attitude changes. The majority of flight time is under Epstein thrust, so you can handle corrections with the main engine.

            For a lot of what we see, capability for a few 180 degree flips is probably enough. In which case, compared to bringing extra thruster propellant, CMGs cost more fuel to haul around.

            It’s similar to the many existing spacecraft that don’t use CMGs. Are they worth it over the alternatives for the planned mission?

            Of course ships with specific purposes (such as military) or that operate in particular environments could get a lot more utility from CMGs. I imagine Tycho Station has some huge CMGs.

  2. This is an excellent summary of the ways in which SA Corey does/does not apply real life physics to The Expanse novels. I think you’re right that one of the key elements the authors have going for them is a relatively faithful adherence to “reality” – at least a 4 to a 4.75 on the Mohs scale of science fiction hardness (

    One area where I wonder if “real” space weapon tech would diverge from the Expanse’s vision is lasers. At least as far as I’ve read (halfway through the 2nd book), there is relatively little reliance on lasers. Conversely, John Lumpkin’s “Through Struggle, The Stars” provides a good illustration of how real life space battles could come down to whoever has the best laser optics and the fastest computers to run them. Imagining a sufficiently strong power source (a fusion engine should suffice) and advanced enough lenses, lasers should serve as both excellent point defense against the Expanse’s ubiquitous torpedoes, and as nearly indefensible long range weapons. I’m curious about the near omission of laser weapons – perhaps they’re omitted to set the works apart from the pew-pew-iness of other science fiction, but it may actually be an inadvertent departure from “realism.”

    What do you think?

    1. Dubyrunning: one of the problems with lasers over long distances is that laser beams will, no matter how well you focused them, diverge. When astronomers bounce lasers off the moon to measure its distance from Earth, the return beam has a Texas-sized spot! Maintaining power over vast scales is going to be quite tricky. Close in – say, tens of kilometers, off the cuff – laser weapons might be effective. For communication they would be totally fine.

      1. I am coming to this WAY late (2 years later!).

        On lasers, I’ve never thought they would be terribly effective weapons. Light is easily deflected or scattered. Simply releasing a cloud of crystals in front of/around your ship should be enough to scatter the beam, no? Add that with reflective armor and heat syncs (which the Expanse universe has some sort of handle on given the stealth tech as described) and lasers become ineffective weapons. Not worth the energy spent.

        1. In actuality lasers are based on one of the four fundamental forces of the universe. Electromagnetism is what holds atoms and molecules together. High intensity lasers have interesting interactions with matter that make them virtually the best weapon imaginable. For example the reflectivity of matter decreases as laser intensity increases; and at certain intensities matter automatically flashes to plasma due to the team’s electric field being strong enough to rip electrons from their orbitals. Beam divergence, and hence effective range, is a function of frequency and aperture size.

          But laser combat in space is not very interesting. The short range of coilguns can make for some drama.

  3. Wow, that’s a very nice summary/review of physics in The Expanse! Just what I was looking for!
    I’m just catching up with the book series and it’s a lot of fun.
    While I can understand many decisions of the authors in favor storywise, I’m missing the idea of a space elevator – would fit perfectly in the fictional universe 🙂

    With all these awesome discoveries in 2015 (new horizons, rosetta missions just to name a few) it’s such a great time for solar system exploration / hard SF (looking at you, (The) Martian!).

    I can also recommend the RPG books of “Eclipse Phase” (cc by) wich take place in our solar system and share a lot of ideas with Coreys’ Expanse universe!

  4. I have been re-reading the books, and Google Now suggested this article for me.

    A few things, based on my re-reading (and I wish it was ‘harder’ Sci-Fi).

    The Epstein drive isn’t just magic – it’s just a more efficient fusion rocket.

    I have not done the maths myself, but plenty of others have, and there is plenty of power in fusion for interplanetary flight at constant thrust, up to and exceeding 1G. The fusion products and/or extra material is the reaction mass and not too much is needed.

    The books don’t go into huge details, but it seems earlier fusion drives turned very little of the released energy into thrust, and had big issues dealing with heat – actual problems fusion rocket designs face. The Epstein drive solves those issues without explaining how, but the whole system is plausible using more advanced tech.

    They never state how much fuel / reaction mass they have, but I noticed one interesting thing. (and I don’t know if this is luck, or researched).

    They often seem to fly around at 1/3rd G, which gives the days to weeks of travel time in between Jupiter or closer planets But at one point, travelling to the Ring out past Uranus, its stated that the trip takes a couple of months.

    It turns out, taking a few months to do that trip is probably 0.05 G or less acceleration. Which actually uses about the same amount of energy (and so fusion fuel) as a 0.3 G trip over shorter distances.

    So the Roci probably has a realistic fuel limit that means they can’t just go very high G all the time.

    Exact engine design aside, the Epstein drive clearly has the power to do a planetary take off and landing too. A high speed aerobraking entry is not needed when you can come down slower using the main drive. Sure it’s inefficient, but perfectly plausible.

    In regards to reaction jets, vs reaction wheels or control moment gyros, I think these fusion ships would tend towards thrusters.

    They have stated a few times that various ships have a teakettle mode, where they use fusion power to heat water and make a steam jet. It’s needed when docking, getting close to other ships etc, so avoid hitting them with any fusion rocket exhaust. For docking to a spinning space station or asteroid, the thrusters would need to be quite powerful – matching Ceres spin for example. A reaction wheel is not helpful for more than attitude control.

    So you have ships that already have a very powerful thruster system that has all the energy it needs via fusion. Sure, reaction mass is limited. But the amount of heading changes that could be done with a reaction wheel (as little as just the mid point flip over) probably does not save enough reaction mass, compared to the cost of fuel and reaction mass to carry it around all the time.

    The thing that bugs me most about the series is the lack of automation. Which I imagine is because otherwise things get super boring, and there is no seat of the pants pilot work needed.

    1. @Lindsay Handmer

      Excellent point about automation!

      While I haven’t done sample calculations, it’s hard to run at any constant acceleration for long without running out of reaction mass – or implausibly high (“magical”) specific impulse. It’s the latter point that I’m picking on.

      You might be surprised just how powerful reaction wheels or control moment gyros are compared to thrusters. You’re right, they don’t provide any maneuver delta-v – only attitude control – but they are still a staple of agile spacecraft and I imagine the Roci would have a powerful set of CMGs on board.

      1. Yeah, specific impulse is the problem – but I am not sure that what we see in The Expanse is implausibly – at least not in terms of being outside the current understanding of physics.

        I don’t know enough to calculate it myself, but other websites (such as Atomic Rockets) suggest that Deuterium Helium 3 fusion direct thrust exhaust velocity is up to 8.9% light speed. It’s still an incredibly powerful fusion reactor / drive, but doesn’t need an excess amount of reaction mass.

        The part ‘solved’ by the Epstein Drive is efficiently turning that fusion torch into usable thrust (and avoiding heat buildup) with a new magnetic field rocket bell design.

        They still can’t fly around at high G forever. But a few AU at 0.3 G, or a months long trip at a fraction of that seems fine.

        Whether or not humans every make a workable fusion drive, something along the lines of the Epstein drive seems to be a reasonable goal for the future.

        But yeah, it is fiction, so there are plenty of issues, or non-issues, depending on how it all works. The thing that bugs me is the Roci’s water tank. It’s the same size as the cargo hold, which while small (and dimensions never given), seems big enough to contain hundreds of tons of water. Maybe that is all for the thrusters (or used as reaction mass in the main drive), but seems overkill.

        I do agree that the Roci would (or should) have CMGs – they lose some thrusters in one fight and can’t maneuver properly, which suggests no CMGs. But being a military ship, it should have both systems of attitude control, for battle redundancy.

        But (and this once again depends on exactly how things like the main drive work), I can’t see the utility of CMGs on a lot of ships. The power is not the issue – it’s how much use they would actually get / vs the cost and weight penalty. While CMGs are a stable of agile spacecraft now, the Expanse ships are a different beast IMO.

        The ISS for example uses CMGs to keep a fixed attitude compared to Earth, so they get used all the time. They have a range of ways to unload CMG momentum without thrusters, so it saves on propellant.

        The Expanse ships need some sort of balance system when the main drive is on, to counteract variances of center of gravity. Probably the magnetic bottle on the main drive can be vectored slightly.

        But the ships wouldn’t need constant attitude control from the CMGs, and even if they did, they still need to use something like thrusters to unload them. Maybe some ships can unload the momentum in CMGs when docked, but then the station still needs to deal with it.

        CMGs would make more sense I think if the ships could use their fusion drives all the time – such as docking, but had limited thrust vectoring. Then you could use the main drive for maneuvering, and the CMGs for attitude control. On long flight, then small amounts of thrust vectoring could be used to unload the CMGs.

        Or say something like the Razorback – however racing works, it may need to make a lot of attitude changes while the main drive is running. Way more than most ships, such as the Cant, which would not turn much. Say that the attitude changes are beyond that vectoring the main drive can do. Then thrusters and reaction mass capable of a lot of attitude changes might be less efficient than CMGs. Because CMGs could help handle the peak attitude changes, but be constantly unloaded by the main drive between fast turns.

        But since the expanse ships need and already have powerful thrusters and don’t need to turn much, outside of specific reasons (such as military ships), I can’t see CMGs adding enough utility to offset the cost and weight penalty.

        Anyway, it all depends how the ships really work, which is not known. It’s fun to speculate though. Makes me want to write my own near future space adventures!

        1. That’s exactly how I feel when I read books like The Expanse – it’s why I write articles like this one, too! Let me know if you do write anything. 🙂

          Atomic Rockets is a good place to look for those calculations. I bet you’re correct in that a fusion drive could run at something like 0.03g for the durations mentioned in the Expanse books, but remember that the books often describe those long hauls as one-third g. That’s more than an order of magnitude bigger, and Tsiolkovsky’s rocket equation isn’t our friend!

          You’re clearly done your basic research on momentum control systems. But I think you’re missing the single biggest advantage CMGs or reaction wheels have: they allow attitude maneuvers without spending any propellant whatsoever. “Running teakettle” would be an extremely mass-inefficient way to get torque on your spacecraft, requiring lots of water to blow out into space, and a modestly sized reaction wheel or CMG could do better while only spending electrical power. You are correct about periodic momentum unloads (I work with some of the guys who developed those maneuvers for ISS, and I programmed the momentum unloading system on an upcoming small satellite!) but even then it’s worth it. Your idea of thrust vectoring would solve about two thirds of that problem right there!

          Thanks for the discussion!

          1. Yeah, future space rocket discussions are my favourite! So I got curious enough to do some of my own calculations(and using various of the calculators linked from Atomic rockets) – which are probably pretty suspect, or I might have missed something, so please confirm if so.

            IIRC, some of the longer hauls we see at 1/3g are a few AU in distance – Tycho Station to Io for example. Of course it depends on all the relative positions, but a 5 AU capability (for two way journeys without refueling) at 0.3 g seems reasonable for what’s in the book.

            So 5 AU at 0.3g takes 280 hours. 1,008,000 seconds at 3.27 m/s acceleration is a Delta V cost of 3,296,160 m/s. So a lot.

            Our theoretical direct fusion exhaust rocket velocity is 8.9% c according to Atomic rockets. So 26,681,528 m/s.

            Using the Tsiolkovsky rocket equation, then the Roci needs to be about 15% fuel / reaction mass to do what we see in the books. Depending on how much she weighs, and how efficient her fusion reaction is, then less than 1% of her mass would be actual fusion fuel.

            That also leaves decent wiggle room to have less efficient propulsion, or lower exhaust speed, or more reaction mass, and still be able to maintain 0.3g for long periods.

            I don’t know if I messed up the numbers, but if accurate, the Expanse ship behaviour seems pretty reasonable to me. Sure, it’s an incredibly powerful fusion reactor, with an unknown magnetic containment / nozzle system. But as far as I can tell, it doesn’t need any magic to bypass known physics, and checks out in terms of reaction mass needed.

            The economics of it would be interesting though. Holden says that fuel costs a small fortune. But doubling the travel time uses about a quarter of the energy, so he must value their time / 0.3g a lot.

  5. I think “valuing” their time is only half the story. First of all, I bet that crew tasks are probably more efficient at 0.3g. Also, 0.3g is probably enough “gravity” to negate (or at least reduce) most of the medial problems with long-term exposure to micro-gravity environments. Remember how important it was for the Nauvoo/Behemoth to get it’s spin-gravity going when people with stranded in the slow-zone.

    If 0.3g seems to be a standard for most commercial transports… then it is probably the sweet-spot in terms of fuel-efficiency vs. crew-health. Matrian ships would feel right at home at 0.3 g. I bet UN Navy ships zip around faster that Martian or Belter ships because military budgets can afford the higher fuel consumption and their crews probably benefit being closer to 1g. That fits the Earther “consumption-stereotype”, too.

    1. The alternative is all ships using spin gravity though, not just coasting in free fall. It adds complexity, but the weight penalty of say a tether system should not be huge.

      So really the economics are fuel costs + extra ship complexity / weight, vs travel time.

      Holden says that fuel is expensive, which is fair enough. But doubling travel time uses a quarter of the energy.

      Sure, speed is always going to be more important in some cases. But I think the decision to use mostly thrust gravity is for a simpler plot, rather than making for a self consistent story universe.

      Which is totally fair enough – but it’s still interesting to speculate on the economics of it all.

  6. I think the Expanse does a pretty good job using the right “gravity” technology in the right spots.

    Spin gravity is best for stationary “stations” or ships that will end up spending a lot of time “on the float” (ex: the Navoo). Constant thrust is best for ships to actually go places.

    As for CMGs… I think once you are talking about volumes of propellant needed for a fusion drive to work in a constant-thrust scenario… the amount needed for occasional attitude changes is negligible. I cannot imagine the mass of a CMG big/powerful enough to maeneuver the Roci would be less than a little bit of extra plumbing needed to get propellant to RCS thrusters.

    I bet any space technology NOT moved by an Epstein drive probably could get the CMG treatment. Satellites, drones, probes. MAYBE missiles? For some reason I like the idea of not needing to “unload” the CMG in a missile because… who cares?

    1. Not to mention, a ship like the Roci will already need powerful RCS thrusters for translation movement with the Epstien drive off (which we see).

      So the only weight penalty is the extra reaction mass, which won’t be much on most missions.

      I like the concept of missiles ‘unloading’ their CMGs into their targets!

      But I think for an internally self consistent universe, spin gravity and constant thrust would actually be used together. I can see why they didn’t go down this route, in terms of the ‘harder’ science and complexity involved distracting from the story, so I am not suggesting they should have done this instead. But I think it’s interesting to consider.

      The Navoo is a special case, because it’s not designed to do both thrust and spin gravity. In contrast, Tycho station is – the spinning sections also pivot.

      According to Holden, fuel is very expensive. Earth to Mars (as an example) takes a 3.2 days (on average) at 0.3g.

      If instead, they fly at 0.03g, the trip takes 10.1 days. But uses 1/10th the energy / fuel, which is a huge difference when fuel is expensive.

      Even dropping back to 0.1g, the trip time is only 5.5 days, but it uses 1/3rd the fuel.

      With expensive fuel, I don’t think thrust gravity is worth it over spin gravity. With a pivoting spin gravity section of the ship (even just on cables), then you just spin a little slower to compensate for thrust. Just instead of spinning straight outwards, it angles towards the engine a little under thrust (like Tycho station), and gravity inside feels the same.

      It also means that you can have gravity for a larger portion of the trip, and vary it up to a full G without having to expend huge amounts of fuel.

      Another ship behavior not explored is the economics of the fuel cost, vs reaction mass cost. Of course we don’t know exactly how an Epstien drive works, but it’s a fusion rocket with a high exhaust velocity, to enable constant thrust without needing huge amounts of reaction mass.

      But if reaction mass is cheap, while fuel is expensive, then trading off exhaust velocity for higher acceleration and more reaction mass usage could be cheaper. And it could vary depending on the mission / ship / economics, but constant thrust might not always be the best bet.

      So keeping the same fusion energy output, more reaction mass is accelerated, but not as fast. That gives higher thrust, so higher acceleration. Instead of accelerating the entire time, you accelerate harder for the start of the trip, and then coast, then decelerate harder at the end.

      (and this is presumably how the Roci etc can accelerate very hard – they use more reaction mass)

      Depending on how it’s varied, you can have the same travel time (or faster travel time), for less fuel use, but more reaction mass use. If reaction mass is cheap and fuel is expensive, then it’s a cheaper way to travel.

      You will need spin gravity in the middle though, and more reaction mass means less cargo space.

      But I would think that for something like the Cant, reaction mass (water) is “free”, so the economics come down to things like fuel use, vs cargo value etc. Of course, hauling ice like the Cant does probably doesn’t make economic sense.

      Anyway, great series, which purposefully avoids turning readers off with exactly these sort of harder sci fi considerations. But it’s fun to speculate!

  7. Anyone notice that Epstein’s space jalopy seems to owe a lot to Robert Heinlein’s Torch Ships? The show had not, that I remember, to that point explained why or how the general solar system ships could have as much fuel as Hopalong Cassidy had bullets in those old B westerns. The physics is not explained and I don’t think the conservation laws of physics allow it.

  8. I love the series but I think the two biggest tech problems are
    1) getting rid of heat (see Saturn Run)
    2) radiation poisoning (and anticancer drugs won’t fix LD50 items

    1. If you are really into tech problems, google the “Atomic Rockets” website.

      That being said… I think that the Expanse strikes the right mix of science and story. It isn’t hard sci-fi… and the authors deliberately did not try to make it so. The actual performance of the Epstein drive is not outside the realm of what physics would allow.

      I think the most unique part of the Expanse is that it is close enough future for us to make decent guesses on what the technology will actually be like…and it explores the under-served story-space of interplanetary (but not interstellar) travel.

      1. I’m actually against this approach towards realism in science fiction. It is not a great evil that has some benefits but poisons the story if followed too closely.

        The entire objective of my blog, for example, is to reconcile the wishes of authors and creators with the science in hard sf instead of standing in their way.

  9. Outstanding! Serious thank you. Everything made sense to me (spinning gravity etc). But I couldn’t get the thrust gravity unless the ships were stacked (skyscraper style as you mentioned it) and I must have missed that part when I was reading.

    Even then, I suspected that the ships would gain their cruising speed and cut engines so I didn’t understand why they would still experience even partial G. The Epstein drive and the acceleration thrust/deceleration thrust makes sense.

  10. Thanks, Joseph. Very informative, and the Expanse really rocks, both print and cable. I loved your analogy of the baseballs and the ice! I know you enjoyed Miller pouring that drink on Ceres in Season 1. Coriolis, indeed!

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.