Cheat Sheet

For some good general notes on designing spacecraft in general, read Rick Robinson's Rocketpunk Manifesto essay on Spaceship Design 101. Also worth reading are Rick's essays on constructing things in space and the price of a spaceship.

For some good general notes on making a fusion powered spacecraft, you might want to read Application of Recommended Design Practices for Conceptual Nuclear Fusion Space Propulsion Systems. There are also some nice examples on the Realistic Designs page.

For less scientifically accurate spacecraft design the Constant Variantions blog has a nice article on historical trends in science fiction spacecraft design.

Like any other living system, the internal operations of a spacecraft can be analyzed with Living Systems Theory, to discover sources of interesting plot complications.


The improvised space warcraft are the type that seems to hold the most story potential.  These would, as mentioned, likely be built by colonies that are in conflict.  As they do not have to operate in an atmosphere, and are built by relatively poor colonies, they are likely to be rather crude.  The basic components required are structure, propulsion, weapons, life support, power, sensors, control, and communications, and each will be briefly discussed in turn.

There are two methods of assembling an improvised warcraft, either adapting an existing vessel, or constructing a new one from parts.  The use of an existing vessel removes the need for some, though not all, of the various components.  The structure will obviously be preserved, and propulsion and life support are almost certain to remain unchanged as well.  Weapons would obviously be added, along with their associated control equipment.  Existing power supplies might or might not be sufficient, depending on the weapons fit chosen.  Sensors and communications are gray areas, depending on the tasks required of the vessel, and the existing fits in these areas.

Structure is one of the easiest components to create.  So long as the builder does not mind the craft being somewhat heavy, slapping a few beams together should be sufficient.  Existing structures, such as cargo containers, could easily be modified, or simple new ones fabricated.  In any case, this is not likely to be a driving factor in construction or conversion of vessels.  Any group incapable of creating basic structure is also almost certainly incapable of surviving if it were to win a rebellion.

Propulsion for improvised craft is likely to be chemfuel due to the fact that it is by far the simplest to implement, and has sufficient delta-V for any operations that do not involve transiting deep space.  It is entirely possible that a colony will have standard chemfuel engines used in various places, and one of them, with appropriate fuel tanks, would be fitted to the vessel.  Nuclear propulsion is much more expensive, and might well involve detailed engineering to avoid killing the crew.  The performance advantage over chemfuel for nuclear-thermal is probably not significant enough to justify the problems involved, and nuclear-electric is only practical for vessels intended for deep space use.  

Weapons are a tricky issue. These are likely to be improvised as well, and would fall into the same categories already discussed.  Improvised lasers are highly unlikely.  Industrial lasers lack the optics trains required for weapon use, while any optical trains available (probably from astronomical or other scientific sources) are unlikely to be able to handle the high powers output by the lasers.  With some time, an appropriate optical train could be designed and mated to an industrial laser, and it is even possible that colonies could design and test such things in case of war.  Kinetic projectors are in largely the same boat.  While small mass drivers could be adapted to such a task, it is difficult to see a role for such devices on a colony.  There is also the issue of targeting, which, while not insoluble, requires good pointing accuracy and possibly the creation of guided projectiles, which have even less peacetime use then the launchers themselves.  

This leaves three options, missiles, lancers, and unguided kinetics.  Unguided kinetics can be as simple as junk thrown out of the airlock, but they are of very limited effectiveness, as shown in Section 8.  Missiles and lancer projectiles face many of the same issues, and the only practical difference is the motor, which should be relatively easy to improvise.  A missile or lancer will require sensors, thrusters, and guidance logic of its own.  As this force is presumably facing another more or less improvised one, complicated guidance logic is probably unnecessary, and proportional navigation is quite easy to implement.  The sensor might well be adapted from another role, which means that the likely problem is in the thrusters.  These must be well-balanced and integrated with the guidance logic.  Depending upon the materials available, this could range from very easy (if there are large numbers of small, self-propelled objects that can be used as warheads lying around) to extremely difficult (if the entire system must be designed from scratch).  Small thrusters themselves are an unknown.  There are some systems that might use small thrusters, such as thruster packs for spacewalkers, in which case the actual integration is all that is required.  However, the number available might well be strictly limited, forcing the builder to start from scratch.  Note that this is not as easy as it seems.  While a primitive kinetic could probably be built with the simplest of systems, it would be inefficient, of dubious reliability, and probably quite large.  In the end, this is an area in which the specific situation plays a very large role, leaving us unable to anticipate exactly what might occur.  

Life support should be straightforward to build into a vessel.  Any space colony will undoubtedly have small, portable habs that can be used for surface expeditions or what have you.  Mounting one of those would be relatively simple, and the actual mechanisms for short-term life support are fairly rudimentary, easing implementation if for some reason a hab had to be constructed from scratch.

Power is a rather tricky proposition.  Unless a nuclear propulsion system is used, power is likely to be at a premium.  Most non-nuclear power studies assume that solar panels will be used, but these have significant drawbacks for space warfare.  The biggest problem is that solar panels are vulnerable to damage from opponent’s lasers or powder weapons, and cannot be angled for protection, unlike radiators.  Radiators, discussed in Section 7, are both somewhat less vulnerable to damage, and can be kept edge-on to the enemy.  A clever opponent could manage to create a dilemma between getting power and preserving the panels from damage.  Alternatives include fuel cells and batteries.  Fuel cells are the current solution for short-duration spacecraft, due to their possessing higher specific energy than batteries.  The problem is that fuel cells are somewhat involved to manufacture and are not likely to be common on space colonies, unless the colony is far enough from the Sun that solar panels are not effective.  Batteries are somewhat more likely to exist, but are heavy for their power output.  The only bright side is that a truly improvised spacecraft is unlikely to need much in the way of power, particularly when compared to a nuclear-electric laserstar.

Sensors are probably the biggest unknown.  A proper space warcraft needs some form of active sensor to localize the enemy, although it is possible that a simple passive sensor would be adequate for simple missiles.  The sensors might or might not be readily available.  Sensor suites for existing spacecraft are the most likely source, although cobbling sensor suites together from other uses might be possible.

Control is mostly a matter of systems integration.  Depending on the nature of the systems involved, control setup could range from simple running of cables and plugging together a few modules, to having to write all of the code to make everything talk, or simply doing without an integrated control system.  While it is certain that some systems will have to talk (sensors and weapons spring to mind) a large portion of the integration could be skipped, with a resulting loss of efficiency due to the crew having to move around.

Communications is fairly simple, as one thing people in space will have to do is talk.  This should ensure the availability of communications modules, which can be attached to the vessel.  The most likely cause of problems is lack of strong encryption and particularly electronic warfare capability in such modules.  Depending upon the capability of the opposition, this may or may not be a serious hindrance.  The encryption capabilities should be a reasonably simple fix, involving mostly software updates.  The EW work will be harder, as there are likely to be physical changes required to ensure freedom from interference by enemy radio transmissions.  This is not likely to be a problem if the communications module is radio-based.

by Byron Coffey (2016)


RocketCat sez

This is the living breathing core of all rocket design. Delta Vee equals Vee Ee times Natural Log of Arr. This is the secret that makes rocket design possible. Now it is time to see the practical application of the key to rocketry.

Everything about fundamental spacecraft design revolves around the Tsiolkovsky rocket equation.

Δv = Ve * ln[R]

The variables are the velocity change required by the mission (Δv or delta-V), the propulsion system's exhaust velocity (Ve), and the spacecraft's mass ratio (R). Remember the mass ratio is the spacecraft's wet mass (mass fully loaded with propellant) divided by the dry mass (mass with empty propellant tanks).

The point is you want as high a delta-V as you can possibly get. The higher the delta-V, the more types of missions the spacecraft will be able to perform. If the delta-V is too low the spacecraft will not be able to perform any useful missions at all.

Looking at the equation, the two obvious ways of increasing the delta-V is to increase the exhaust velocity or increase the mass ratio. Or both. Turns out there are two more sneaky ways of dealing with the problem which we will get to in a moment.

Historically, the first approach has been increasing the exhaust velocity by inventing more and more powerful rocket engines. Unfortunately for the anti-nuclear people, chemical propulsion exhaust velocity has pretty much hit the theoretical maximum. The only way to increase exhaust velocity is by using rockets powered by nuclear energy or by power sources even more frightful and ecologically unsound. And you ain't gonna be able to run a large thrust ion-drive with solar cells.

The second approach is increasing the mass ratio by reducing the spacecraft's dry mass. This is the source of the rule below Every Gram Counts. Remember that the dry mass includes a spacecraft's structure, propellant tankage, lifesystem, crewmembers, consumables (food, water, and air), hydroponics tanks, cargo, atomic missiles, toilet paper, clothing, space suits, dental floss, kitty litter for the ship's cat, the ship's cat itself, and other ship systems. Everything that is not propellant, in other words. All of it will have to be trimmed.

To reduce dry mass: use lightweight titanium instead of heavy steel, shave all structural members as thin as possible while also using lightening holes, make the propellant tanks little more than foil balloons, use inflatable structures, make the floors open mesh gratings instead of solid sheets, hire short and skinny astronauts, use life support systems that recycle, impose draconian limits on the mass each crewperson is allowed for personal items, and so on. Other tricks include using Beamed Power so that the spacecraft does not carry the mass of an on-board power plant, and avoiding the mass of a habitat module by hitching a ride on an Aldrin Cycler. Finally the effective mass ratio can be increased by multi-staging but that should be reserved for when you are really desperate.

The third approach is trying to reduce the delta-V required by the mission. Use Hohmann minimum energy orbits. If the destination planet has an atmosphere, use aerobraking instead of delta-V. Get more delta-V for free by exploiting the Oberth Effect, that is, do your burns while very close to a planet. Instead of paying delta-V for shifting the spacecraft's trajectory or velocity, use gravitational slingshots. NASA uses all of these techniques heavily.

If your technology is high enough, use space tethers, launch catapults, and MagBeams.

The fourth and most extreme approach is to cheat the equation itself, to make the entire equation not relevant to the spacecraft. The equation assumes that the spacecraft is carrying all the propellant needed for the mission, this can be bent several ways. Use Sail Propulsion which does not use propellant at all. Use propellant depots and in-situ resource utilization to refuel in mid-mission. The extreme case of ISRU is the Bussard Ramjet which scoops up propellant from the thin interstellar medium, but that only works past the speed of 1% lightspeed or so.

In our Polaris example, given the mass ratio of 3, we know that the Polaris is 66% propellant and 33% everything else. Give the total mass of 1188.9 tons means 792.6 tons of propellant and 396.3 tons of everything else. Since each GC engine is 30 tons, that means 150 tons of engine and 246.3 of everything else.

Every Gram Counts

RocketCat sez

Listen up, rocket designers. Write these words in letters of fire on your cerebellum. Every Gram Counts! Add an extra gram and you will pay for it with extra propellant as if the Mafia loan shark wants you to pay up with liquid hydrogen. With interest compounded hourly.

The only exception is if you are dealing with an Orion drive spaceship or other torchship.

The most fundamental constraint on designing a rocket-propelled vehicle is Every Gram Counts.

Why? Short answer: This is a consequence of the equation for delta-V.

Why? Slightly longer answer: As a general rule, a rocket with the highest delta-V capacity is going to need three kilograms of propellant for every kilogram of rocket+payload. The lower the total kilograms of rocket+payload, the lower the propellant mass required. This relates to the second strategy of rocket design mentioned above.

Why? Long Answer:

Say the mission needs 5 km/s of delta-V. Each kilogram of payload requires propellant to give it 5 km/s.

But that propellant has mass as well. The propellant needed for that original kilogram of payload will require a second slug of propellant so that it too can be delta-Ved to 5 km/s.

And the second slug of propellant has mass as well, so you'll need a third slug of propellant for the second slug of propellant — you see how it gets expensive fast. So you want to minimize the payload mass as much as possible or you will be paying through the nose with propellant.

This is called The Tyranny of the Rocket Equation.

Even worse, for a given propulsion system, the easiest way to increase the delta-V you can get out of that system is by increasing the mass ratio. It probably is not economical to push the mass ratio above 4.0, which translates into 3 kg of propellant for every 1 kg of rocket+payload. And it is nearly impossible to push the mass ratio above 20. Translation: spacecraft with a mass ratio of 20 or above are basically constructed out of gossamer and soap bubbles.

This is why rocket designers are always looking for ways to conserve mass.

Orion drive spacecraft and other torchships are not subject to this constraint, because they are unreasonably powerful.

It also does not apply to "stationary" items such as space stations and planetary bases, since they do not move under rocket propulsion. In fact, the added mass might be useful to stablize a space station's orbit, or as additional radiation shielding. Rocket vehicles might use aluminium, titanium, magnesium, or other lightweight metal as their structural material; but a space station would be better off using heavy iron or Invar.

The only consideration is if the station or base components have to be transported to the desired site by a rocket-propelled transport. Then it makes sense to make the components low mass, because then the station bits are payload. It makes even more sense to construct the space station or base on site using in-situ resources, so you don't have to eat the transport costs at all.

Everything Is Connected

Like aircraft and sea-going warship design, one soon discovers that everything is connected to everything else. When the designer changes one aspect of the design this causes a series of related changes to ripple through the rest of the design.

For instance, if the designer reduces the propellant tank capacity by 5% this has implications for the spacecraft's mass ratio. If it is important for the spacecraft's delta V to stay the same, the payload will have to be reduced by the same amount. This might cut into the amount of life support consumables carried, which will reduce the number of days a mission can last. If the same amount of scientific observations have to be done in the reduced time, another crew member might have to be added. This will decrease the mass available for consumables even more. And so on.

The technical term is "cascading changes." The only thing worse is cascading failures.

As an example, I have some notes of graphing various space warship optimizations with examples of how various warship classes map onto the graph.


(ed note: Captain Randall, his crew, and his submarine were inadvertently put into suspended animation in 1945. They have been awakened one thousand years later, and must go into battle with alien invaders. Captain Randall is talking over the refurbishment of his submarine with a robot named Austin. The discussion is about limitations in submarine design, but the "everything is connected to everything" principle still holds true.)

     "I'll take your word for it, Austin. At the moment there are more important things. What is the state of the ship?"
     "As on the day of trials, sir, without fault, and completely refitted. The experts have many ideas for improvements, sir, which will be submitted to you in due course."
     "And the munitions?"
     "None, sir. They had deteriorated beyond safe use. However, they can be replaced or, if you prefer it, something more violent substituted."
     "One thing at a time. At the moment, have them replaced. Does this age understand explosives?"
     "Chemical history does, sir. Your ammunition holds can be filled in approximately half an hour and that includes the torpedo tubes. Will you continue with ancient compressed air, underwater missiles, sir? We have some smaller and far more efficient."
     "Normal torpedoes with a normal warhead, please, Austin."
     "As you wish, sir."
     Randall smiled faintly. "You sound disappointed, Austin. So, since you have a logical mind, I will be logical. You probably have in mind more advanced and far more destructive missiles which came into being just before man decided on peace for all time. No doubt these missiles are faster, have a greater range, are self-seeking and are several times more destructive. How big are they, Austin?"
     "About the quarter of the size of yours sir. I estimate you can carry—"
     "They won't fit our torpedo tubes, will they? This will mean interior adaptation, rebuilding, all of which will have to be related to buoyancy, speed—above and below surface—diving and surfacing. Which method is going to be the quickest?"
     "I had overlooked that angle, sir." There was considerable respect in the robot's voice.
     "That is because you are not yet a sailor—and let me assure you. You cannot become a seaman from text books or even convenient memory banks: It's a question of experience. No doubt, Ordinary Seaman Austin, your mind is brimming with ideas for replacing the ship's motors with something modern, smaller, a hundred times more efficient and one tenth the weight. This would play hell with our diving, upset our displacement and require the alteration of our tank capacity to meet our increased buoyancy. Am I making myself plain?"
     "Too plain, sir. With respect, I am learning a great deal."
     "Good. Don't misunderstand me, Austin. I want your ideas and I shall ask for them, but only when you, yourself, can relate them to experience."

From THE TIME MERCENARIES by Philip E. High (1968)

Fundamental Design

As mentioned in Rick Robinson's Spaceship Design 101, all spacecraft are composed of two sections: the Propulsion Bus and the Payload Section.

The Propulsion Bus has the propulsion system, propellant tankage, fuel container (if any), power plant, power plant heat radiator (if any), anti-radiation shadow shield (if any), and a keel-structure to hold it all together. Sometimes the keel is reduced to just a thrust-frame on top of the engine, with the other components stacked on top.

The Payload Section is what the propulsion bus is pushing from planet to planet. It can include crew, flight control station, propulsion/power plant control station and maintenance center, astrogation station, detection and communication equipment, habitat module with life support equipment (including environmental heat radiators) and consumables (air, food, water), space taxis, space pods, and docking ports.

But most importantly, the payload section must contain the reason for the spacecraft's existence. This might be organized as a discrete mission module, or it might be several components mounted around the payload section.

Planetary Exploration Vessel
Scientific instruments, space ferries, airless landers
Cargo Vessel
Cargo holds
Extra propellant tanks (and a remarkably large mass ratio)
Space Tug
Ship grappling equipment, push plates, and an over-sized high-thrust propulsion bus.
Coast Guard vessel
rescue equipment, ship grappling gear, ship repair supplies space taxis, space pods, and a propulsion bus with extra delta V.
Orbital Guard vessel
Telescopes and other tracking gear, nuclear detonation detectors, asteroid redirecting equipment, weapons
Troop transports
Huge habitat modules
Blockade Runner
Stealth technology, small cargo hold, and a propulsion bus with high acceleration

You get the idea.

A warship's payload section can include anti-spacecraft weapons, orbital bombardment weapons (for revolt suppression type spacecraft as well), weapon mounts, weapon control stations, combat information center, armor, point defense, weapon heat radiators and heat sinks, and anything else that can be used to mission-kill enemy spacecraft.

Pirate ships and privateers might forgo defenses if they only expect to be engaging unarmed cargo ships. But they will regret this if they have the misfortune to encounter armed enemy convoy escort ships or are surprised by a Q-ship.


I will argue that deep space craft have essentially two sections that can largely be treated separately from one another. One section is the propulsion bus — drive engine, reactor if any, solar wings or radiator fins, propellant tankage, and a keel structure to hold it all together. The other is the payload section that it pushes along from world to world.

There are both conceptual and economic reasons to treat them separately. Conceptually, because a propulsion bus might push many different payloads for different missions, such as light payloads on fast orbits versus heavy payloads on slow orbits. A little noticed but important feature of deep space craft is that you cannot overload them. They do not sink, or crash at the end of the runway, or even bottom out their suspension. They merely perform more sluggishly, with reduced acceleration and (for a given propellant supply) less delta v.

Conceptual logic is also economic logic. The outfits that build drive buses would like to sell them to lots of different customers for a broad range of assignments.

This is not necessarily an argument for true modular construction, with drive buses hitching up to payloads on an ad hoc basis like big-rig trucks and trailers. Building things to couple and uncouple adds complexity, mass, and cost — plug connectors, docking collars, and so forth. Moreover, drive buses intended for manned ships need to be human-rated, not just with higher safety factors but provision for supplying housekeeping power to the hab, etc. But these things, along with differing sizes or number of propellant tanks, and so forth, can all be minor variations in a drive bus design family.

The payload we are most interested in is, naturally, us. The main habitat section of a deep space ship closely resembles a space station. It is likely that habs intended for prolonged missions will be spun, for health, efficiency, and all round convenience. (Flush!) The design of a spin hab is dominated by the spin structure and — unless you spin the entire ship — the coupling between the spin and nonspin sections.

Because ships' spin habs have the features of stations they may be used as stations, and again we can imagine design families, with some variants intended for ships and others as orbital platforms having only stationkeeping propulsion. Habs are the one major part of a deep space ship that correspond fairly well to our concept of a hull. Spin habs are entirely different in shape, but the shape is constrained; once you build it you can't easily modify it, beyond adding another complete spin section.

For those with bank cards at the ready, buying a deep space ship might be not unlike buying a computer. If your mission needs are fairly standard, you check off options on a menu. Those with more specialized requirements can select major components — perhaps a drive bus from one manufacturer, a main crew hab from another, along with custom payload sections, service bays, and so forth, assembled to your specifications.

In fact, both technology and probable historical development suggest that fabrication and overall assembly will be two distinct phases, carried on in different places, quite unlike either shipyard or aircraft assembly practice. In the early days, large deep space craft will be built the way the ISS was, assembled on orbit out of modules built on Earth and launched as payloads. In time fabrication may move to the Moon, or wherever else, but final assembly (at least of larger craft) will continue to be done at orbital facilities. I call them cageworks, on the assumption that a cage or cradle structure provides handy anchoring points for equipment.

For game or sim purposes, my advice would be to treat drive buses and hab sections as the primary building blocks for ships, whether these components are permanently attached to each other or simply coupled together. Both approaches might be in use.

A couple of provisos. All of the above applies mainly to deep space craft, especially with high specific impulse drives. Ships for landing on airless planets have some similar features. Ships that use rapid aerobraking, however, are aerospace craft and broadly resemble airplanes, even if they never land or even go below orbital speed.

And I have said nothing of warcraft. Kinetics are essentially just another payload. Lasers, and other energy weapons such as coilguns, probably draw power from the drive reactor, calling for some modifications in the drive bus. These things don't much affect the overall configuration. Armor protection would, but discussions here have left me doubtful of its value against either lasers or kinetics. Laser stars and other major warcraft may not be dramatically different in appearance from civil craft of similar size.

Nyrath: I'm reminded of Sir Arthur C. Clarke's early space science books. He noted that a nuclear powered spacecraft would probably resemble a dumb bell, that is, two spheres connected by a stick. The hab module is the forward sphere, the nuclear drive is the rear sphere, the stick is long to provide some inverse square protection from radiation, and the propellant tanks would be on the stick, probably clustered near the nuclear drive.

One can also imagine modules designed by diverse corporations being incompatible with others on purpose. "Not invented here" syndrome.

One can also imagine a tramp freighter composed of incompatible modules, being held together with bailing wire and spit.

Qwert: The compatibility between modules will mostly depend on how the market develops. One extreme maybe Microsoft, a monopoly that basically sets the standard, as everything has to adhere to it. The other one extreme is the current hardware industry. Your memories and components have to fit on every motherboard if you want to sell them.

On the field of big aircraft manufacturing, standardisation dominates almost everything... excerpt the end product. If you manufacture engines, they have to fit on Airbus as well as on Boeing's aircraft. The components industry is dispersed and competition is intense.

On the other side, a pilot trained to flight with a Boeing can´t immediately switch to an Airbus without some training. Competition centers around two big players and nobody is interested in making life easier to the other.

In short: how spaceship components will be build, will mostly depend on how the industry evolves. A monopoly, strict government regulation, competition between many small producers or a highly dispersed specialised components industry may benefit a system of standards. On the other side competition between a small number of giants, may produce different incompatible systems.

Rick Robinson: True modularity is by no means a given. But some features of modularity, call it demi-modularity, are inherent to deep space technology.

You probably want to keep your propellant tanks separate from the corrosive, explosive stuff we breathe. Drive engines are essentially bolted onto the tail. Generally the major parts of a deep space ship don't have to fit together snugly. If you want to hang something out on a bracket you probably can.

Ferrell: I think that the engine package, mated to a suitable tank, mated to a hab module, mated to a mission module (a seperate entity from the hab module) would (as the last step in design) incorporate the heat management system suitable to the final design. Then construction/assymbly would occure.

If there aren't landers/shuttles at the final destination and you need them, then you can carry landers/shuttles in place of cargo. A mission module may be an extended docking module that a number of small modules 'plug' into, (or your transfer craft).

Manuvering thrusters should be at mutliple points/modules (distributed from the 'nose' to the 'tail').

Of course, you could stand it all on it's ear and have the mission module be on the inside of the ship, the hab ring be around the middle (with its radiators in arcs between its connecting pylons),with the engines, tanks, powerplants, radiators, and nav sensors clustered around both ends of the mission module; and any docking would be at the tips of the mission module, or on the inside walls of the mission module.

Nyrath: There might be a brisk trade in "interface modules", that would connect modules made by different manufacturers. I'm reminded of the Apollo ASTP Docking Module used in the Apollo-Soyuz mission. This was a tiny airlock module with a NASA style docking collar on one end, and a Soviet style docking collar on the other.

Not to mention the International Space Station Pressurized mating adapters.

Rick Robinson: Sabersonic — Yes, I'm gliding over a host of devils in the details. The payload section will surely have attitude thrusters, for example, and these must coordinate with attitude thrusters on the drive bus end.

Nyrath — 'Standards,' indeed! Again this is a promo for full service cageworks that will provide things like docking adapters.

Jean Remy: On standards:

Currently the main type of freight vessel is the container-carrier. They are favored because the containers can be loaded/unloaded easily, then simply popped on a freight train or a truck. I can see the same things for a cargo spaceship. There won't be a cargo "module" but rather anchoring hard points for standard containers. Those containers will be loaded on a booster to orbit, transferred onto a ship, then at the other end the container is loaded into a simple remote-controlled lifting body (for planets with an atmosphere, say, Mars) or just a simple frame with thrusters (for the Moon) where they will be loaded into maglev trains if needed, etc.

The command post/bridge:

I'd still put in in the non-rotational part of the ship, certainly on warships (you want it as deep inside the ship as you can for protection) Not only that, but I would stop rotation in combat: precession might alter your maneuverability, and damage might weaken the structure or cause a wobble which would rip your ship apart. I would also put it in the middle for a civilian vessel, as then the command module can double as storm cellar, or if there's a separate storm cellar, you'll still have access to the command post during a solar flare. It's also a good defense against micrometeorites. The command center is just too important to risk placing it on the outer rim of a ship, no matter how convenient or comfortable it would be to the crew.

Sabersonic: Even so, it would probably be prudent to not have all command, or at least navigation control, be monopolized by the CIC. At minimum and barring mass budgeting, there should be two for overall spacecraft control: One that is the CIC/Navigation primary with the second being the engine room for mostly emergency purposes. If only because for something as complex and (for the first few decades if not centuries of interplanetary travel, let alone interstellar) inevitably as fragile as a spacecraft, one should avoid "putting ones eggs into one basket" and to always "have a plan B". Space and interplanetary/interstellar travel is not a place for the ill-prudent, and that's just the natural dangers.

As for standards, well, before standards could ever become "standard" for lack of a better word, there lies the inevitable "Format War" in one form or another that would occur when one believes that their system is more efficient and reliable than the other or worse: a new industry/business standard that has the potential to supplant the jobs of numerous Dock workers or in this case "Cage Workers" that could drive hostility and perhaps a little bit of political pressure before the whole matter is settled in one way or another. It would not be a full parallel to the Teamsters Strike that is basically Teamsters vs Trucks, but it would not be a completely quiet deal when paychecks are involved.

Rick Robinson: My division of ships into drive buses and payload sections is more because of operational factors than manufacturing considerations.

For example, ships inherently have (at least) two big 'hull' structures, the crew hab and the main propellant tank. These are probably at very different temperatures, which right away is a big reason to keep them physically separate.

Using the propellant tankage as shielding is appealing, but even with a vacuum separation it means that your 290 K hab shell is dumping 350 W/m2 of waste heat right into your 20 K liquid hydrogen tanks.

Jean — The containership principle seems highly likely for most space cargo, with the standard pod being defined originally to fit Earth orbital shuttle bays.

In this case, your typical space freighter is a drive bus pushing a rack structure with clamps for pods. The rack might be configurable so that you can also carry 'oversize' loads.

Jean and Sabersonic — In a parallel discussion at SFConsim-l, the question was raised whether civil ships need a 'control room' at all, or whether people could just stand watch from their regular work stations.

I think that any spacecraft with a fair number of passengers or crew will have a watch at the main life support panel, because life support is always running, has constantly changing loads, and things can go very bad very quickly. The life support panel will almost certainly be in the spin section, because that is where the life support is.

You may as well put the engineering panel here as well. There's no reason for an 'engine room' in the maritime sense, since the drive is mounted externally, and if it is nuclear you don't want to go anywhere near it.

No doubt you could maneuver and navigate the ship from here as well. But en route there is very little of this to do.

The only time you really conn a (civil) spacecraft is during rendezvous and docking, or similar evolutions. At these times you surely de-spin, but you might want a separate control station next to the main airlock, with viewports for maximum situational awareness.

Because of its location, this station would also naturally serve as the ceremonial 'quarterdeck' where VIPs are greeted, and ordinary mortals report aboard.

Warcraft are a whole 'nother matter, with protective considerations arguing for a control room at the center of the hab section.

Jean Remy: Multiple Command posts: Yes. That's why I used the term command post rather than CIC, Bridge, DCC, Main Engineering. It was meant to be generic and refer to all these. There ought to be at least CIC and DCC, if not a redundant third post, all of them capable of doing every job, but each ideally suited to one. The ship is more efficient with all of them, but can still work with the loss of one or two.

Decentralized Command posts: Possibly, but I don't think it likely. First of all we could do that now, but we don't. Psychologically, I think crews want to have the "Captain on Bridge" so to speak. It is generally recognized that good officers are the ones who stick by their men when the going gets tough. Caesar rode into battle at Alesia, rallying his troops to victory when it looked uncertain. Washington rode with his men. Patton led from a tank. Now granted the captain of a ship can't go anywhere, but seeing him on the Bridge will be a boost to morale.

Rick Robinson: The question of having a control room at all was in the context of civil spacecraft. If they have an sudden emergency it is most likely to be a life support crisis such as fire, for which the classic 'bridge' functions are fairly irrelevant.

Military craft are a special case, and I'd certainly expect them to have a control center.

And maybe all ships, because to extend a point Jean makes, existing space programs are quasi-military in origin. The military outlook toward emergency response is coded into their DNA, so to speak.

All human carrying deep space ships will need a storm shelter in any case, and it would be fairly natural to configure this as an emergency control center.

Luke Campbell: Re: Shielding. It now seems likely that a plasma magnet generated by a low mass antenna could deflect any charged solar radiation, so the crew would be safe from flares and CMEs. It does not seem like a plasma magnet could stop galactic cosmic rays, GCRs are a steady source of background radiation, not the sort of thing that a "storm cellar" would help with. A plasma magnet also would not protect against neutral particle radiation, but the only neutral particle radiation likely to be a threat is man-made: neutron radiation from nukes and possibly high energy photons from x-ray or gamma ray lasers.

For those not in the know, a plasma magnet uses low frequency radio waves to produce a rotating field that induces a current in the surrounding solar wind plasma. This current forms a dipole magnetic field that deflects and reflects charged particles. The field is not strong enough to deflect solar wind protons, but it does deflect the electrons, leading to charge separation that pulls the protons back to the electron cloud before they reach the section being protected.

Yes I was trying to stop GCRs (but calculations showed it would need an unreasonably huge magnet). The plasma magnets wouldn't stop the solar wind protons either, when considered as individual particles — you need the plasma effects of the electrons to stop the solar wind. This lets you get by with a much smaller magnet.

However, there is one possible additional method of mitigating the GCR dose - medication. As we learn more about cellular repair and cell "suicide", new treatments may become possible for both chronic and acute radiation poisoning (and oddly, you are likely to want the opposite reaction in these two cases - for chronic exposure, you want the damaged cells to destroy themselves to prevent cancer; for acute exposure you want the damaged cells to repair themselves to prevent anemia, hemophilia, a compromised immune system, and digestive difficulties). Incidentally, it has been shown that vitamin D helps with chronic radiation exposure, although the mechanism is not clear.

Rick Robinson: To keep cryogenic propellant from boiling off on long missions you will need active refrigeration, pumping heat out of the tank. Otherwise heat buildup will be inexorable.

The ship will be designed to keep the propellant tanks away from the main radiator fins and such, and generally minimize heat absorption from the rest of the ship, so the heat you mainly have to deal with is from sunlight.

The tanks will be painted white or silvery to reflect away most sunlight. I assume that you can reflect about 90 percent. For a spherical tank at 1 AU, that means about 35 W/m2 of absorbed solar radiation that you'll have to pump out of the tank.

A 20 meter diameter tank holds about 250 tons of hydrogen, or 1500 tons of methane. Surface area is 1250 m2, so at 1 AU you'd need 44 kW of refrigerating capacity, i.e. heat extraction, to keep propellant cold.

At Jupiter distance, 5 AU, solar flux is reduced by 96 percent, and you only need 2 kW of refrigeration for hydrogen - none for methane, which will stay liquid or even tend to freeze.

Suppose you put a 10 meter diameter hab inside the tank {to give the hab some radiation protection}. (This only reduces tank capacity by 1/8.) But even with a vacuum layer you will have IR heating from the hab surface, at room temperature: 400 W/m2 * 314 m2 = ~125 kW.

So in this case putting the hab inside the fuel tank multiplies your propellant refrigeration bill by 4x. Which is a lot, but not horrible; the shielding might be worth it. But wrapping propellant around a spin hab is tougher.

Rick Robinson: Modular design is always a tradeoff. You get more operational flexibility, at cost of more complicated/heavier/weaker connections. Integral designs will be favored when the components will consistently be used together.

Much will depend on tech. Torch type drives and even 'conventional' nuke electric drives pretty much have to be mounted on a pylon, which sort of invites the option of unbolting it from the rest. OTOH, as you note, the drive section may well have its own control center. And since the rest of the ship sits on top of the pylon, it's a fine line between 'pylon' and 'chassis.'

On naming, I could also make a case that the crew hab compartment is the main component, and so would be named. Especially if it is a spin gravity structure. And 'spaceships' may end up having more than one name, just as a named train might included Pullman cars with names of their own.

And if ships are highly modular, some terms might be borrowed from railroading. For example, 'consist' as a noun (pronounced CON-sist) for the whole assemblage. Thus, 'The Ty Cobb departed Mars with a consist of [such and such modules].'

Amusing side note: modular spacecraft reverse the order of trains: the 'locomotive' or drive engine is at the back (more precisely the base), while the 'caboose' or control cabin might well be at the front/top.

From SPACESHIP DESIGN 101 by Rick Robinson (2009)

This section is intended to address some gaps in available information about spacecraft design in the Plausible Mid-Future (PMF), with an eye towards space warfare.  It is not a summary of such information, most of which can be found at Atomic Rockets.  The largest gap in current practice comes in the preliminary design phase.  A normal method used is to specify the fully-loaded mass of a vessel, and then work out the amounts required for remass, tanks, engine, and so on, and then figure out the payload (habitat, weapons, sensors, cargo, and so on) from there.  While there are times this is appropriate engineering practice (notably if you’re launching the spacecraft from Earth and have a fixed launch mass), in the majority of cases the payload mass should be the starting point.  The following equation can be used for such calculations:

Where P is the payload mass (any fixed masses, such as habitats, weapons, sensors, etc.), M is the loaded (wet) mass, R is the mass ratio of the rocket, T is the tank fraction (or any mass that scales with reaction mass) as a decimal ratio of such mass (e.g., 0.1 for 10% of remass), and E is any mass that scales with the overall mass of the ship, such as engines or structure, also as a decimal.  

This equation adequately describes a basic spacecraft with a single propulsion system.  It is possible to use the same equation to calculate the mass of a spacecraft with two separate propulsion systems.

The terms in this equation are identical to those in the equation above, with R1 and T1 representing the mass ratio and tank fraction for the (arbitrary) first engine, and R2 and T2 likewise for the second.  Calculate both mass ratios based on the fully-loaded spacecraft.  If both mass ratios approach 2, then the bottom of the equation will come out negative, and the spacecraft obviously cannot be built as specified.  Note that when doing delta-V calculations to get the mass ratio, each engine is assumed to expend all of its delta-V while the tanks for the other engine are still full.  In reality, the spacecraft will have more delta-V than those calculations would indicate, but solving properly for a more realistic and complicated mission profile requires numerical methods outside the scope of this paper.

One design problem that is commonly raised is the matter of artificial gravity.  In the setting under discussion, this can only be achieved by spin.  The details of this are available elsewhere, but these schemes essentially boil down to either spinning the entire spacecraft or just spinning the hab itself.  Both create significant design problems.  Spinning the spacecraft involves rating all systems for operations both in free fall and under spin, including tanks, thrusters, and plumbing.  The loads imposed by spin are likely to be significantly larger than any thrust loads, which drives up structural mass significantly.  This can be minimized by keeping things close to the spin axis, but that is likely to stretch the ship, which imposes its own structural penalties.  A spinning hab has to be connected to the rest of the spacecraft, which is not a trivial engineering problem.  The connection will have to be low-friction, transmit thrust loads, and pass power, fluids, and quite possibly people as well.  And it must work 24/5 for months.  All of this trouble with artificial gravity is required to avoid catastrophic health problems on arrival.  However, there is a potential alternative.  Medical science might someday be able to prevent the negative effects of Zero-G on the body, making the life of the spacecraft designer much easier.  

When this conclusion was put before Rob Herrick, an epidemiologist, he did not think it was feasible.

     “The problem is that they [the degenerative effects of zero-G] are the result of mechanical unloading and natural physiological processes. The muscles don't work as hard, and so they atrophy. The bones don't carry the same dynamic loads, so they demineralize. Both are the result of normal physiological processes whereby the body adapts to the environment, only expending what energy is necessary. The only way to treat that pharmacologically is to block those natural processes, and that opens up a really bad can of worms. All kinds of transporters would have to be knocked out, you'd have to monkey with the natural muscle processes, and God knows what else. Essentially, you're talking about chemically overriding lots of homeostasis mechanisms, and we have no idea if said overrides are reversible, or what the consequences of that would be in other tissues. My bet is bad to worse. As the whole field of endocrine disruptors is discovering, messing with natural hormonal processes is very very dangerous.

     Even if it worked with no off-target effects, you'd have major issues. Body development would be all kinds of screwed up, so it's not something you'd want to do for children or young adults. Since peak bone mass is not accrued until early twenties, a lot of your recruits would be in a window where they're supposed to still be growing, and you're chemically blocking that. Similarly, would you have issues with obesity? If your musculature is not functioning normally (to prevent atrophy), how will that effect the body's energy balance? What other bodily processes that are interconnected will be effected? Then you get into all the effects of going back into a gravity well. Would you come off the drugs (and thus require a washout period before you go downside, and a ramp-up period before you could go topside again)?

     Spin and gravity is an engineering headache, but a solvable one. Pharmacologically altering the body to prevent the loss of muscle and bone mass that the body seems surplus to requirements has all kinds of unknowns, off target-effects and unintended consequences. You're going to put people at severe risk for medical complications, some of which could be lifelong or even lethal.”

This is a compelling case that it is not possible to treat the effects of zero-G medically.  However, if for story reasons a workaround is needed, medical treatment is no less plausible than many devices used even in relatively hard Sci-Fi.

The task of designing spacecraft for a sci-fi setting is complicated by the need to find out all the things that need to be included, and get numbers for them.  The author has created a spreadsheet to automate this task, including an editable sheet of constants to allow the user to customize it to his needs.  The numbers there are the author’s best guess for Mid-PMF settings, but too complicated to duplicate here.

Rick Robinson’s general rule is that spacecraft will (in the sort of setting examined here) become broadly comparable to jetliners in cost, at about $1 million/ton in current dollars.  This is probably fairly accurate for civilian vessels, at least to a factor of 3 or so.  Warships are likely to be more expensive, as most of the components that separate warships from civilian ships are very expensive for their mass.  In aircraft terms, an F-16 is approximately $2 million/ton, as is the F-15, while the F/A-18E/F Super Hornet is closer to $4 million/ton.  This is certainly a better approximation than the difference between warships and cargo ships, as spacecraft and aircraft both have relatively expensive structures and engines, unlike naval vessels, where by far the most expensive component of a warship is its electronics.  For example, the ships of the Arleigh Burke-class of destroyers seem to be averaging between $150,000 and $250,000/ton, while various cargo ships seem to hover between $1000 and $5000/ton.

As mentioned in Section 5, some have suggested that the drive would be modular, with the front end of the ship (containing weapons, crew, cargo, and the like) built separately and attached for various missions.  This is somewhat plausible in a commercial context, but has serious problems in a military one.  However, the idea of buying a separate drive and payload and mating them together is quite likely, and could see military and civilian vessels sharing drive types.  (This is not as strange as present experience would lead us to believe.  It was only during WWII that military aircraft clearly separated from civilian ones in terms of performance and technology.)  This simplifies design of spacecraft significantly, as one can first design the engine, and then build payloads around it.

One common problem during the discussion of spacecraft design is the rating of the spacecraft.  With other vehicles, we have fairly simple specifications, such as maximum speed, range, and payload capacity.  However, none of these strictly applies in space, and the fact that spacecraft are not limited by gravity and movement through a fluid medium makes specifying the equivalents rather difficult.  Acceleration and delta-V obviously depend on the masses of the various components, which can be changed far more readily than on terrestrial vehicles, and cargo capacity is limited only by how long you’re willing to take to get where you’re going.  A replacement might be a series of standard trajectories, and the payload a craft can carry on them.  This works well if all of the spacecraft being rated are generally similar in terms of performance, and take similar trajectories in reality.  However, it does not work as well in a scenario where different types of ships take wildly different trajectories with different amounts of cargo.  In that scenario, ships might be rated by the minimum time for certain transfers (Earth-Mars at optimum, for instance) with a specified payload, either a fixed percentage of dry mass, or a series of specified masses for various sizes of ships.  This allows a comparison between ships of different classes, but within a class (liner, bulk cargo, etc.) the first method would probably be preferred.

A related problem is the selection of an appropriate delta-V during preliminary design.  In some cases, this is relatively easy, such as when a spacecraft is intended to use Hohmann or Hohmann-like trajectories, as numbers for such are easily available.  But such numbers are inadequate for a warship, or for any ship that operates in a much higher delta-V band, and unless the vessel has so much delta-V and such high acceleration that Brachistochrone approximations become accurate (and even then, if the vessel is not using a reactionless drive, the loss of remass can throw such numbers off significantly, unless much more complicated methods are used, numerical or otherwise).  The author has attempted to fill this gap by creating a series of tables of delta-Vs and transit times between various bodies, with the tables giving the percentage of the time that a vessel leaving one body can reach another within a specified amount of time with a given amount of delta-V.  The tables can be found at the end of this section.  Table 9 covers Earth-Mars transits, while Table 10 describes Earth-Jupiter transits.

The tables are generated in MATLAB by solving Lambert’s problem for a large number of departure days and transit times, and calculating the delta-V to go from stationary relative to the departure planet to stationary relative to the destination planet.  This involved assuming that there was a single instant delta-V burn at each end, which is a good approximation if the burn time is short compared to the transit time, as it would be for chemical or most fission-thermal rockets.  For systems which burn a significant amount of the time, this approximation is not as good, and the tables should only be used as a general guide to the required delta-V.  

Each table is the composite of 16 different tables generated with different starting geometry, and with each table containing data from at least one synodic period.  Note that this was all done in a sun-centric system, and that the delta-V necessary to deal with either planet’s gravity well was not included.  This will add some extra delta-V, the necessary amount shrinking in absolute terms as the overall delta-V is increased due to the Oberth effect.  The decision to not include escape and capture delta-V was made because to do otherwise would have involved specifying reference orbits to escape from and capture to, and would have added significant complexity to the program at a minimal gain in utility for most users.  

One thing that is apparent from these tables is the degree to which Jupiter missions are more hit-or-miss than Mars missions.  For Jupiter transfers, 84% of the options are either going to be viable all of the time, or not going to be viable at all.  For Mars, the equivalent value is only 56%.  Some of this is due to the much larger and more variable time increments used in the Jupiter calculations, but much of it is due to the fact that the geometry changes significantly less between Earth and Jupiter than it does between Earth and Mars.

It should also be noted that these tables are an attempt to find an average over all possible relative positions of the two bodies.  For the design of an actual spacecraft, analysis would instead start with modeling of geometries over the projected life of the spacecraft.  The approximations given here are reasonably close for theoretical use, but should not be used to plan actual space missions.  

Heat management is a vital part of the design and operation of a space vessel, particularly a warcraft.  Section 3 mentioned some of the issues with regards to stealth, but a more comprehensive analysis is necessary.  There are two options for dealing with waste heat in battle: radiators and heat sinks.  If the waste heat is not dealt with, it would rapidly fry the ship and crew.  

All space vessels will need radiators to disperse the heat they produce as part of normal operations.  If using an electric drive, power (and therefore waste heat) production will be no higher in battle then during cruise.  This would allow the standard radiators to be used indefinitely during battle without requiring additional cooling systems.  The problem with radiators is that they are relatively large and vulnerable to damage.  The best solution is to keep them edge-on to the enemy, and probably armor the front edge.  The problem with this solution is that the vessel is constrained in maneuver, and can only face one (or possibly two) enemy forces at once without exposing the radiator.  If the techlevel is high enough to make maneuver in combat a viable proposition, then radiators are of dubious utility in combat.  On the other hand, the traditional laserstar battle suits radiators quite well.  The faceplate and the forward edge of the radiators are always pointed at the enemy, and almost all maneuvers are made side-to-side to dodge kinetics.  The only problem is vulnerability to a direct kinetic hit.  If a projectile were to arrive precisely edge-on, it could tear the entire radiator in two.  Bending the radiator slightly would eliminate this vulnerability, but would also increase armor requirements.  However, even a bent radiator would still have issues with grazing impacts.  A projectile coming in very close to parallel with the radiator’s surface would tend to tear a long hole in it, as opposed to the small hole left by a projectile traveling perpendicular to the surface.  However, the low-incidence projectile would have to penetrate much more material, so kinetics designed for such attacks would naturally have more mass or fewer pieces of shrapnel than one designed for normal attacks.

Heat sinks avoid the vulnerability to damage of radiators, but have a drawback of their own.  By their very nature, they have a limited heat capacity, which places a limit on how much power a ship can produce during an engagement, and thus on the duration of an engagement.  If the heat sinks fill up, the ship would begin to fry unless the radiators were extended immediately.  In the game Attack Vector: Tactical, extending the radiators is used to signal surrender.  Obviously, the heat clock is a major disadvantage, but it is necessary when the vessel expects to expose several aspects to the enemy.  

One topic that briefly needs to be addressed is electric propulsion. In discussions, VASIMR-type engines are usually considered the baseline.  However, Dr. Joshua Rovey of Missouri S&T told the author that Hall Effect thrusters today are capable of the sort of performance that VASIMR is currently promising after development is finished.  VASIMR is apparently getting attention due to good marketing people.  

Another topic that deserves discussion is the effect of nuclear power on spacecraft design.  For large warcraft, nuclear power, both for propulsion and for electricity is a must-have.  Even if the design of solar panels advances to the point at which they become a viable alternative for providing electrical propulsion power in large civilian spacecraft, there are several major drawbacks for military service.  The largest is that solar panels only work when facing the sun, unlike radiators, which work best when not facing the sun.  The distinction between the two is important, as it is nearly always possible to find an orientation which keeps the radiator edge-on to the enemy and still operating efficiently, while a solar panel must be pointed in a single direction, potentially exposing it to hostile fire.  A solar panel is particularly vulnerable to laser fire, as it is by nature an optical device.  While hard numbers on this are surprisingly difficult to find, it appears that damage will probably occur to photovoltaics when exposed to intensities of around 300 KJ/m2, for short pulses (<10-4 seconds), with threshold requirements increasing from there as the pulse length increases.  For a CW laser (>1 second), the power flux required for damage is approximately 10 MW/m2.  Photovoltaics can also be attacked using small particles such as sand, as described in Section 7 for use against lasers.  While a full analysis of the potential damage is beyond the scope of this section, it appears that sand would be a reasonably effective means of attacking photovoltaics, particularly given the large area involved.  The size of a solar array also complicates maneuvering the panels edge-on to the incoming particles, and could potentially raise structural concerns.

Radiators, on the other hand, are more resistant to damage.  Firing lasers at them will only decrease the thermal efficiency of the reactor slightly, as the radiator is designed to disperse heat.  Particle clouds that are designed for surface effects would be ineffective against a properly-designed radiator, or at very best reduce the emissivity by a small amount.  Small pieces of shrapnel designed to pierce the radiator entirely would be the best means of attack (described above), as fully armoring a radiator is likely to be impossible because of the mass requirements.  

However, it could be argued that this ignores the vulnerability of the reactor itself to damage.  While in Hollywood, “They’ve hit the reactor!” is usually followed by a massive explosion, that is not the case in reality.  First, the reactor is a very small target, usually shielded by the bulk of the ship, so it’s unlikely to be hit in the first place.  Second, nuclear reactors simply do not turn into bombs under any circumstances, and particularly not random damage to the core.  The few cases in history in which a reactor has gone prompt critical (SL-1 and Chernobyl being the best-known) were caused by poor procedure, and are vanishingly unlikely to happen due to random damage.  

That said, it still seems a potentially bad idea to put all of one’s eggs in a single basket.  Solar panels are highly redundant, but the reactor could still be put out of action with a single hit.  The response to this is fairly simple.  First, there are reactor designs using heat pipes that have sufficient redundancy to continue operation even if the reactor core itself is hit.  The specific heat pipe will be put off line, but if the design has 150, that’s not a great worry.  Second, the reactor and associated gear (power converters and such) are buried deep in the ship, where they will be difficult to get at, and the power converters can be duplicated for redundancy.

One last concern is the ejection of reactor core material after a hit, and the potential for said material to irradiate the crew. This is also probably minor, as the crew is still being shot at, and the spacecraft will have some shielding against both background radiation and nuclear weapons. (Thanks to Dr. Jeffrey King of the Colorado School of Mines for providing much of the material on space nuclear power and propulsion.)

A couple of other issues with nuclear power are relevant and of interest.  The first is the choice of remass in nuclear-thermal rockets.  While hydrogen is obviously the best possible choice (the reasons for this are outside the scope of this paper, but the details are easy to find), it is also hard to find in many places.  With other forms of remass, the NTR does not compete terribly well with chemical rockets, but it can theoretically use any form of remass available.  The biggest problem with alternative remasses is material limits.  With most proposed materials, oxidizing remasses will rapidly erode and destroy the engine.  The alternatives to avoid this are rhenium and iridium, which are both very expensive, explaining why they are not in use today.  However, both elements are common in asteroids, making them viable choices in a setting with large-scale space industry.

As discussed in Section 7, vibration is a serious issue for laser-armed spacecraft.  Any rotating part will produce vibrations, and minimizing these vibrations is of interest to the designer.  While there is undoubted a significant amount that could be done to reduce the vibrations produced by conventional machinery (the exact techniques are probably classified, as their primary application is in submarine silencing), it seems simpler to use systems with no moving parts, which should theoretically minimize both vibration and maintenance.  Heat pipes, as mentioned above, are an entirely passive means of moving heat around, both from a reactor to an energy converter (which could mean a turbine, a thermocouple, or any of the other wonderful things engineers can think of) and from the energy converter to the radiator.  There are also electromagnetic pumps for liquid metal which while not entirely passive, but will cut down on the vibration load.

There are even proposed systems of energy conversion which are reasonably efficient and involve no moving parts.  The best-known of these proposed systems is probably the Alkali Metal Thermal-to-Electric Converter (AMTEC), which has been extensively studied.  However, a recent effort by NASA to bring the technology into deployment failed, giving the technology a bad name.  There are some, however, who believe it still holds promise.

If systems like AMTEC are not available, the spacecraft will have to use conventional hat engines.  These are likely to use one of the two standard thermodynamic cycles, the Brayton cycle (gas turbine), and the Rankine cycle (steam turbine).  The primary difference between the two is that in the Brayton cycle, the working fluid remains a gas throughout, while in the Rankine cycle, it moves from liquid to gas and back again.  In theoretical design, radiators are normally sized assuming constant temperature throughout, which is true for most Rankine cycle systems (as the radiators are where the fluid condenses at a constant temperature) and produces the well-known result that radiator area is minimized when the radiator temperature is 75% of the generation temperature.  However, this is not true for Brayton cycle radiators.  There is no convenient mechanism to release the necessary heat at a constant temperature, so the radiator performs differently as the gas cools.  There is not a simple formula here, but an iterative procedure can be used to minimize radiator area for an ideal system (which is close enough for our purposes).  Use of this method does require some knowledge of the basics of gas turbine propulsion, but it is not terribly esoteric. (Thanks to Dr. David Riggins of Missouri S&T for presenting this material in class. Those who have had more experience in propulsion and fluid dynamics might recognize some simplifications of the explanatory material, and some nomenclature changes. This was intended to hold down the length of this section and clarify it without sacrificing accuracy of the results.)

An ideal gas turbine can be thought of as being made of 4 separate stages.  First, isentropic compression, which means that there is no heat transfer and all energy put into the system by the compressor, is used to compress the gas instead of heating it.  Second, isobaric (constant-pressure) heat addition, which occurs in the reactor.  Third, isentropic expansion through a turbine, which outputs mechanical work (the goal of this whole process).  Lastly, isobaric heat rejection, through the radiator, which returns the working fluid to the condition it was at before entering the compressor.  The compressor and turbine are defined by their pressure ratios, written as πc and πt respectively.  The pressure ratio is the pressure of the fluid after the component divided by the pressure ahead of the component.

For this method, values for Cp, γ, ηc, ηt, and T3 must be selected.  Cp is the constant-pressure specific heat capacity of the fluid, while γ is the ratio of specific heats.  Definitions and values for various fluids can be found online.  ηc and ηt are the efficiencies for the compressor and turbine respectively.  Values between 0.8 and 0.9 are probably reasonable.  T3 is the temperature at the outlet of the heat addition stage, and is normally set by the design of the reactor itself.  Representative values might be 1600-1700 K for a conventional nuclear reactor, although higher values are possible.  (All temperature values throughout should be in Kelvin, not Celsius.)  

The first step is to select a value for πc, with anything from 2 to 10 being plausible.  Because it is a closed system, πt will be equal to 1/ πc.  Once this is known, it is possible to calculate T4 (temperature downstream of the turbine) using .

At this point, a value for T1 must also be selected.  This is the temperature at the entrance to the compressor.  Using this, the value for T2, the temperature at the compressor exit, can be calculated using .  This allows the overall efficiency of the power-generation system, &eta (work output/heat input), to be calculated using .  All of this information can then be used to find the radiator area per unit work output (A’, m2/W) with where σ is the Stefan-Boltzmann constant (5.670373×10-8) and ε is the emissivity of the radiator (0.9-1.0). Once you have this value, select a different value of T1, and repeat the rest of the paragraph.  When a minimum has been found, select a different value for πc and repeat the entire procedure until a global minimum has been found.  It would probably be a good idea to use a spreadsheet to automate this.

Table 9: Earth-Mars Transits
Maximum Transit Time (Must Arrive No Later Than, Days)
Table 10: Earth-Jupiter Transits
Maximum Transit Time (Must Arrive No Later Than, Days)
by Byron Coffey (2016)

Well, I was going to post the second part of The Shipping Trade today, except that writing it didn’t happen because of day job, and so forth. Then, I thought I might post a sketch of the ship involved, just to give y’all an idea of what you’ll be looking at, but then that would require me to go out and hire a scanner. That, and I made said sketch, and then looked at it, and then concluded that I couldn’t possibly inflict such a terrible picture on my readers…

So permit me, please, instead to sketch a verbal picture for you of the

CMS Greed and Mass-Energy

To start with, Greed and Mass-Energy is atypically large for a free trader; in those leagues, which principally deal in small, high-value-to-mass/volume cargoes, lugging around 40,000 tons displacement of cargo is huge. (It’s still not in the major freight line league, though; those guys can use freighters that are million ton-displacement behemoths.) Thus, the shipcorp that owns her (it’s essentially a syndicate of officers, crew, and former crew, with executive power vested in the captain-owner) is pretty prosperous to be able to cover her running costs. Dealing in brokered cargo actually isn’t her main business – she specializes in contracts like the RCS-assembly charter from Kerbol to Kythera she just left, but an empty hold is a hole that drinks money, so you take the cargo when you can get it.

Also, obviously, at a size like that, she’s not streamlined, or built to land planetside (gravity wells being acutely expensive); and is even rather more massy than anything that most stations like to have dock directly to them. Her cargo’s generally ferried to station, or upwell and downwell, by local lighters at each end of the trip. Rather, she’s built very much in the classic mode; a long, relatively thin, open-frame truss structure. Attached to that, going from fore to aft, we find these different sections of the ship:

Right at the bow, sitting on the end of the main truss, is the command capsule, an ellipsoid slightly stretched along the ship’s main axis, relatively tiny compared to the rest of the ship, and containing, for starters, the bridge and associated avionics systems. (The bridge is actually buried in the center of the capsule, for its protection; it’s displaced off to the front end of the ship, however, because the command capsule is also where the primary sensors are housed to keep them out of the way of cargo, fuel, and drive radiation, and this positioning cuts down on sensor lag. It’s still pretty safe; it’s not like anyone’s going to be shooting at them.) The first of the other two notable features it houses is docks and locks, right for’ard on the axis where it’s easiest to match thrust and spin, which usually houses a couple of cutters used for taking the crew ashore and for occasional maintenance, and a skimmer for in-field refueling. (The fuel itself doesn’t pass through here – the skimmer docks aft to offload what it scoops. No fuel for’ard of the support plate, that’s the general rule.) The second, aft by the truss, is the robot hotel for all the little space-rated utility spiders you may see now and them crawling about the structure doing maintenance, thus saving the engineering department any need to get suited up and go outside for routine work, although they still may need to do so from time to time.

Just aft of that, accommodations and secondary systems are housed in a toroidal gravity wheel. This is actually a very unusual design feature in an Imperial ship-class; just about everyone and especially the spacer-clades are genetically adapted to microgravity, and the spacer-clades prefer it, as a rule; but the Cheneos-class architects originally designed her class for near-frontier work, and included this for occasional passenger service. Greed and Mass-Energy only rarely carries passengers, so they keep it geared all the way down, producing only a tenth of a standard gravity, which doesn’t offend the spacer-clades all that much. There’s a second, smaller wheel rotating inside it to null out the gyroscopic effects; it’s used to house some other equipment that likes a little gravity, but for the most part, this one’s just a countermass.

(The wheel does, however, provide enough gravity to let the CELSS Manager run a pretty decent microbrewery in the spare volume, and perhaps more importantly, provides a place where you can drink it off-shift without suffering from a nasty case of the zero-g bloat. [Remember, folks, bubbles don’t rise in microgravity!] And apart from crew morale, having decent beer makes for good PR when traders meet.)

These areas, incidentally, are one of the few places on board where the really high-tech ontotechnological stuff makes an appearance, in the form of inertial damping. The people who built her liked microgravity, and weren’t all that keen on losing that while under thrust, especially since she was built to fly brachistochrones or near-brachistochrones (bulk tankers and ore freighters, etc., are usually built to fly economic minimum-delta/Hohmann transfers; no-one else wants to wait that long for their cargo) and so would be spending most of her time under thrust. The job of the inertial dampers is to apply the thrust of the drives evenly across the entire area’s structure and everything in it, thus ensuring that no-one actually feels any acceleration, and the lovely microgravity environment is preserved. (It also avoids having to come up with some wretchedly complicated gimbal arrangement for the already wretchedly complicated seals-and-bearings for the gravity wheel, no longer having to do which is something that made architects particularly grateful for this innovation.)

Behind this, the cargo. ‘Way back along the truss there is a very large, solid plate, the support plate. The cargo containers are simply stacked “atop” – by which we mean for’ard – of it, in six big blocks arranged around the axis with sixfold symmetry (this arrangement being a reasonable compromise between use-of-volume and convenient straight lines), and are designed to lock to the plate, the truss, and each other to form a solid interlocked structure. There’s no hold or other walls around the cargo; the containers are themselves spacetight when they need to be, and so lighters can just drop them into place and pick them up freely while in port.

The breakbulk cargo, on the other hand, is messy. It has to be podded up individually when not spacetight, and then individually lashed down and made secure atop the cargo container stacks. This annoys the cargomaster, which is why breakbulk is unpopular these days despite the fact that breakbulk shippers usually pay a premium in exchange for you having to do this (the “lash comp”). Actually, what really annoys the cargomaster is that she can punch a button and have the ship automatically query the v-tags on the container cargo for its mass stats, and so forth, whereas for breakbulk she’s got to recall her Academy training, dig out the spreadsheets, and work out the corrections to the center-of-mass-and-moment-of-inertia chart by hand. Well, still by computer, but you know what I mean.

Aft of the support plate, still in sixfold symmetry, you have the bunkerage – fuel tanks, stacked three deep in multiple rows, all filled with slush deuterium, running right to the stern, where they surround the cylindrical shroud of the mostly-unpressurized engineering hull (you can take a crawlway right back along the truss to the small, pressurized maneuvering room back this far, should you need to examine the drives close-up in flight, but the actual machinery space isn’t), which contains the interlinked systems of the main power reactors and the fusion torches themselves, strapped to the aftmost extent of the main truss.

And there are lots of fuel tanks. Even though said fusion torches are miracles of a mature nuclear technology, capable of achieving near-theoretical efficiencies and outputs and delta-v per unit fuel that routinely makes naval architects from less advanced civilizations throw down their slide rules in despair and weep into their terrible coffee-equivalents, the one unchangeable rule of space travel is that your mass ratio is always much, much less favorable than you might want it to be.

Good thing deuterium’s so cheap, isn’t it?

…and most prominently of all from a distance – dominating the entire view of the ship from a distance, by area as well as by temperature – sweeping out from among the fuel tanks (although comfortably retracted to sit alongside them, leaving approximately a sixth of their radiative area useful, while idling in dock – the vast panels and pipework of the heat radiators. Because the other one unchangeable rule of space travel is that you always have waste heat, too damn much waste heat, and you’ve got to get rid of it somehow. Especially once you fire up those fusion torches. (The radiators, however, unlike the rest of the ship, have only fourfold symmetry – so that they can be perpendicular to each other when unfolded, because there’s very little point in radiating heat right back at your own radiators.)

Rockets Are Not Hotels

RocketCat sez

See that space fantasy at the top? Yep, the good ol' Starship Enterprise. There are two glaring thing wrong with it, right off the bat.

First off, the direction of "down" is almost but not quite totally FUBAR. We'll get into that later.

But secondly, which of the other two ships look most like Enterprise? In terms of blue pressurized habitat module. Yep, the freaking Queen Mary, an ocean liner. Not the Lewis design, a nuclear thermal rocket spacecraft created for a Mars mission. I hope you see the problem.

If you look at most blueprints for the various iterations of the Starship Enterprise, you will notice that every single part of the spacecraft interior is pressurized, with doors, rooms, and toilets. The corridors are wide enough for five people to walk abreast on nice carpeted floors with indirect lighting.

This is ludicrously wrong. And it is not just Star Trek that does this, pretty much all of media science fiction has ships like this. TV Tropes calls this fallacy "Starship Luxurious".

This is an extension of the "Rockets are Boats" fallacy. Passenger aircraft and luxury liners have their entire interior pressurized because so is everything else at sea level on a planet with a breathable atmosphere. For free. So careless starship designers, without a thought, made the unconscious assumption that spacecraft would be totally pressurized as well.

Wrong. Tain't no air in space, and atmosphere is expensive when you have to cart it up out of Terra's gravity well. Not to mention the expensive pressurized hull that has to encase it.

And it is not just the cost of hauling it up the gravity well, the spacecraft's engine has to accelerate the mass of all that junk. Every Gram Counts, so every gram of carpeting, atmosphere, and pressure hull is one less gram of payload, i.e., the reason the spacecraft was created in the first place. See The Tyranny of the Rocket Equation.

In the real world, spacecraft will be mostly tanks of propellant, propulsion system, payload bays, and a lacy lattice-work of support struts holding everything together. The part the people live in will be a tiny pressurized habitat module tucked away somewhere.

Ignorant starship designers have the unconscious assumption that the important part of a spacecraft is the crew, so they designed ships with their priorities reversed. Their ships were mostly gigantic habitat modules with a tiny engine stuck to the rear. Their ships are also ludicrously wrong. If the designers thought about it at all, they might grudgingly include a tiny fuel tank. Which is like the cherry on top of their big icecream sundae of Fail.

So quit drawing ship blueprints with every square inch pressurized and human-accessible. On a real spacecraft if the ship's engineer has to repair the propulsion system, heat radiators, power plant, propellant tanks, or anything like that, they will have to put on their space suit. They will not have the luxury enjoyed by Scotty the engineer, waltzing down a carpeted floor in a shirt-sleeve atmosphere.


The Enterprise's corridors seemed awfully roomy, they were about twice as wide as they should have been. In fact, the whole ship was too roomy. Space is at a premium in any kind of enclosed environment. Anyone who's ever been aboard a submarine—or even an aircraft carrier, for that matter—knows that they are designed for the maximum utilization of their volume. Efficiency is a necessity, and a spaceship is going to have to be designed the same way.

In fact, the requirements of a spaceship are much more stringent—for instance, the interior atmosphere must be maintained with the correct combination of gases, at the right temperature, pressure, humidity, and ionization, to maintain not just the lives, but the comfort as well, of the crew. The margins for deviation are narrow; therefore, every cubic inch of interior volume means airspace that must be maintained—and maintenance requires the expenditure of energy. When you have to conserve your ship's power, you don't waste airspace.

The reason for such broad corridors? They had to be wide enough for a camera dolly, cables and a film crew.

To attempt to show that the ship was cramped would have required the construction of cramped sets—which are harder to work with and would have meant much more in the way of production time.

(So, instead, we're told that the ship has power to waste—it's implied, not specifically said. But if that's so, then we should never see a story in which maintenance of life-support functions become a critical factor for building suspense. A ship can't be both wasteful and limited.)

Another example: the turbo-elevators. These were the machines that took the various crew members from one part of the ship to another. Cood idea; especially as we are told that the thickest part of the Enterprise‘s disc is twelve stories thick.

But—the elevators seemed to be the only way to get from one deck to another. If the ship's power supply were cut off, every deck would be separated from every other. Oh, well, not really—somebody could always crawl through the air vents, or through the Jefferies Tube, or down one of the ladders which we saw very infrequently. All except for the bridge. Cut off the turbo-elevators and you isolate the bridge. Tsk. That's bad designing. Illogical.

Another one: the Captain's cabin. Or anybody’s cabin for that matter. They were all redresses of the same set. If any of those cabins had a bathroom, it was never shown. There weren't even any doors to imply a bathroom. We were never shown the cleanup facilities on the whole ship—not even the sick bay. And the Enterprise was on a five-year mission—isn't that a long time to hold it? Isn't that carrying it a bit too far?

Also, about those cabins—all of the major officers aboard the ship had their own cabins, and roomy places they were too. No complaint here, but what about the crew's quarters? Those were never even shown or suggested. Did each member of the crew have a cabin too? That would have made the Enterprise more of a hotel than a starship. Or did they have bunkrooms?

If they did have bunkrooms, how come they were never mentioned or shown? How come we never got into the crew's lives?

Or was the crew just a collection of some 400-odd androids to walk up and down the halls—scenery behind the main characters, to be moved around as necessary, but not really important to the story except as another part of the background to support the overall illusion?

From THE WORLD OF STAR TREK by by David Gerrold (1973)

The ships above are tail-sitters, so properly avoid the "wrong way is down" problem. But the artist made the second problem much worse. Apparently they figured the entire interior of the spacecraft was for habitable volume. Notice what they got wrong? Well, where the heck is the space for the rocket engines? I lay the blame for this at the artist, I know from experience that writer Jack Williamson knew better.

Modular Construction

An attractive notion is the practice of constructing one's spacecraft out of mix-and-match replaceable components. So if your spacecraft needs to do a planetary landing you can swap the low thrust ion drive for a high thrust chemical rocket. In Charles Sheffield's The MacAndrews Chronicles, the protagonist just calls her ship "the assembly", customized out of whatever modules it needs for the current mission contract.

For detailed examples, see the Boeing Space Tug, NASA Space Tug, JPL Modular Hab System, and the Minimal Volume Spacecraft Cabin.

This will also make ships basically immortal. It will also make it really easy for space pirates to fence their captured prize ships. All they have to do is get the prize ship to the spacecraft equivalent to an automobile chop-shop. There the ship vanishes as an entity, becoming an inventory of laundered easily sold anonymous ship modules with the serial numbers filed off.

One can also imagine junker spacecraft, lashed together out of salvaged and/or junk-heap spacecraft modules by stone-broke would-be ship captains down on their luck.

Or mechanically inclined teenagers who want a ship. This would be much like teens in the United States back in the 1960's used to assemble automobiles out of parts scavenged from the junkyard, since they could not afford to purchase a new or used car. Such teens would gain incredible practical skills as spacecraft mechanics. I wonder if this is how Kaylee from Firefly learned her trade.

Yet another scenario is Our Hero stranded in the interplanetary Sargasso Sea of lost spacecraft, trying to scavenge enough working modules from three broken spacecraft in order to make one working spacecraft.

Rick Robinson notes that attractive as the concept is, there are some practical drawbacks to extreme modularity:

Rick Robinson: This is not necessarily an argument for true modular construction, with drive buses hitching up to payloads on an ad hoc basis like big-rig trucks and trailers. Building things to couple and uncouple adds complexity, mass, and cost — plug connectors, docking collars, and so forth. Moreover, drive buses intended for manned ships need to be human-rated, not just with higher safety factors but provision for supplying housekeeping power to the hab, etc. But these things, along with differing sizes or number of propellant tanks, and so forth, can all be minor variations in a drive bus design family.

Rick Robinson: True modularity is by no means a given. But some features of modularity, call it demi-modularity, are inherent to deep space technology.

You probably want to keep your propellant tanks separate from the corrosive, explosive stuff we breathe. Drive engines are essentially bolted onto the tail. Generally the major parts of a deep space ship don't have to fit together snugly. If you want to hang something out on a bracket you probably can.

Nyrath (me): There might be a brisk trade in "interface modules", that would connect modules made by different manufacturers. I'm reminded of the Apollo ASTP Docking Module used in the Apollo-Soyuz mission. This was a tiny airlock module with a NASA style docking collar on one end, and a Soviet style docking collar on the other.

Not to mention the International Space Station Pressurized mating adapters.

Rick Robinson: Modular design is always a tradeoff. You get more operational flexibility, at cost of more complicated/heavier/weaker connections. Integral designs will be favored when the components will consistently be used together.

Much will depend on tech. Torch type drives and even 'conventional' nuke electric drives pretty much have to be mounted on a pylon, which sort of invites the option of unbolting it from the rest. OTOH, as you note, the drive section may well have its own control center. And since the rest of the ship sits on top of the pylon, it's a fine line between 'pylon' and 'chassis.'

On naming, I could also make a case that the crew hab compartment is the main component, and so would be named. Especially if it is a spin gravity structure. And 'spaceships' may end up having more than one name, just as a named train might included Pullman cars with names of their own.

And if ships are highly modular, some terms might be borrowed from railroading. For example, 'consist' as a noun (pronounced CON-sist) for the whole assemblage. Thus, 'The Ty Cobb departed Mars with a consist of [such and such modules].'

Amusing side note: modular spacecraft reverse the order of trains: the 'locomotive' or drive engine is at the back (more precisely the base), while the 'caboose' or control cabin might well be at the front/top.

From SPACESHIP DESIGN 101 by Rick Robinson (2009)

The ship had several “holds,” actually just enormous, detachable cylinders adapted to carry cargo or passengers. Some of these were sealed and Shaw was reluctant to reveal what was in them. For an unabashed smuggler, that suggested to Thor that some things were unacceptable, even in the freewheeling society of the space settlers. Drive, holds and control were all in separate modules, connected by struts and passage tunnels. It was a common system for ships never intended to make planetfall, allowing great flexibility of size and function. “Also,” Shaw told Thor with a sharklike grin, “it makes it very difficult to keep up with how many and what type of ships are out here. If the authorities were looking for Spartacus, I'd break her up and rearrange her modules with other ships. You can have as many ships as you have command and drive modules.”

It must be a nightmare for customs authorities,” Thor observed.

“We do our humble best. Hijacked ships are never found again because they’re broken up and utilized or sold off as modules. You’ll have to go to a ship sale some time. There’s no pirate hangout like in the holos. Word just gets passed that there’s going to be ship hardware for sale and everybody just sort of congregates at a certain set of coordinates that all the bartenders seem to know about. I've seen whole government military vessels broken up and sold, weaponry and all.”

“Military!” Thor said, aghast. “I thought that was supposed to be impossible. Are there hijackers powerful enough to attack a Space Service ship?”

“Who attacks?” Shaw said. “Usually, it’s just a matter of paying someone to look the other way. The degree of corruption in the higher echelons of the military is immense and has increased tremendously in the last fifteen years. It was historically inevitable. I’ll let you read my monograph on the subject. There are other ways that service vessels make it onto the black market. Sometimes, a whole crew will decide to take early retirement from the service and bring their ship along with them.”

“I think that society out there will be quite different from what I anticipated,” Thor mused.

“I can guarantee it,” Shaw said.

From THE ISLAND WORLDS by Erick Kotani and John Maddox Roberts (1987)

Here's a fat-ass merchant ship that prefers to hand off its cargo in orbit but of course needs to land now and then. It would be a standard hull in Cepheus Engine or MgT First Edition or a partially streamlined hull in CT. It has no business landing anywhere without a beacon, landing lights, and a level field. But of course sometimes it has to. To facilitate docking the original ship has a docking module stuck on the top.

Whether you have rockets, reactionless drives (boo!) or warp drive handwaving docking is a maneuver that gives pilots grey hairs early on. The docking module bears the brunt of it. It's easier and faster to replace a module than an airlock that requires welding a heat resistant hull.

Big companies usually stocked a few modules at starports so if their merchant ships called and were in a hurry or banged their module up they could just swap modules and be on their way to make that hot delivery by the contract deadline. Swap modules and jump the hell out.

As an aside the module also held 10 dTons of cargo for really fast transfer of priority deliveries. If you were lucky the shipment was throw pillows. If you were unlucky munitions but hey no pressure there.

Then some ship builder realized there were a bunch of these 'kegs' just dumped in starports awaiting repair or surplussed. He bought one, stuck a drive and power plant on it and voila! a 20 dTon launch!

The damned things proved popular. They could aid in docking maneuvers since their little bitty engines had more fractional control than the big ship engines. But wait there's more!

The 'caps' at both ends on the keg came off easily (for repairs ... not while docking, that would be a bad thing masquerading as a design feature).

People began modding the modules.

In the 'standard' configuration this keg has the top module fitted with proper landing legs. It docks "nose in" the the ship and of course she can't be used to dock with this feature. If you expect to land her on rough terrain, say in the case of a delivery to a mining outpost on a rocky moon, they provide more stability than those stubby docking clamps.

More modules will be forthcoming. Suggestions will be entertained. I'm pretty sure there's a market for a fuel scoop version with a streamlining module (I know it breaks the at-the-time-of-construction rule but I feel it's justified.) There will be other modules with waldos and power tools for salvage and mining.

(ed note: you might find some inspiration in the modification kits for the Boeing space tug and the NASA space tug)

From SORRY BUT ... by Rob Garitta (2016)

Ship of Theseus Paradox

I thought of a problem with modular designs, based on the ancient Ship of Theseus paradox.

This is my Grandfather's ax.
This is my Grandfather's ax.
My Father replaced the handle.
I replaced the ax-head.
This is my Grandfather's ax.

Is it really still Grandfather's ax or not?

Plutarch first wrote about the paradox in 75 CE. But it was that 17th-century smart-ass Thomas Hobbes who slipped the exploding cigar into the box. He asked the question: what if somebody saves the original discarded handle and ax-head, then assembled them into a second ax. Which one of the two axes is Grandfather's ax? Both, neither, the new one, the old one?

This sounds academic, until you apply it to modular spacecraft.

For purposes of insurance, liability, national registration, contract penalties, mortgages, and a host of other expensive issues; it is crucially important to know the identity of the spacecraft in question. Which ship exactly is being referred to in all those legal documents?

But what if the SS SkyTrash's modules are replaced and the old modules used to make a new ship? Legally which one is the SkyTrash? For that matter, intentionally making a stolen ship vanish by passing it through a spaceship chop-shop can make another set of legal headaches.

The problem of spacecraft identity has got to be legally nailed down.

Don't look to the Theseus Paradox for a solution. The problem was stated almost two thousands years ago and they are still arguing about it

Off-hand I'm not sure what a fool-proof solution would be. My first thought was to attach the identity of the spacecraft to some sine qua non "must-have" ship module. Unfortunately there does not seem to be any. Not all ships are manned, so the habitat module won't work. The only must-have module I see is the propulsion bus (otherwise you have a space station, not a spacecraft). However Captain Affenpinscher might find it strange that the identity of her ship has changed just because she swapped out the propulsion module.

I had a discussion on Google Plus with some of my brain-trust:

Winchell Chung (me): I was mulling over modular spacecraft design, when I suddenly realized I had a "Ship of Theseus" paradox on my hands. Does anybody have any bright ideas about where the legal identity of a spacecraft resides in?

Ray McVay: Keel? Drive core? The route, like railroad trains? In Black Desert, it would be with the AI (spacecraft's installed artificial intelligence)... Annabelle Li (a ship AI) has been two kinds of Heinlein and a CASSTOR, but kept her name. Possibly valid inspection certs for the given configuration is the legal identity of a ship... After all, no transit authority will let one boost in a franken-rocket without giving it a once-over...

John Reiher: I had a thought. There's one component that never changes on any ship. It can be updated, but it can't be replaced, otherwise the Insurance companies will void your policies.

What is it?

The Vehicle Identification Number (VIN) Box. It's a transponder with a unique ID and call sign. It can be customized at time of purchase, but it is the "ship". Anything attached to it becomes the ship.

No VIN Box, no ship.

Two or more VIN Boxes on a single ship, you're breaking the law and you have to inactivate all but one of them.

This is something the Banking and Insurance companies would come up with. They need something that can be unique to each ship, and nothing is more unique than a government issued, sealed, black box VIN Box.

William Black: Modular freighters in my future history setting the Command Module (CMOD) is what the freighter captain actually owns — in a single owner operator context, a transportation company being a single individual or corporate entity that owns many individual CMOD's. However I like +John Reiher's suggestion that the VIN box be the source of identity, and +Raymond McVay's suggestion that identity is locked to the ships AI has obvious merit especially in regards to spacecraft-as-characters. I may steal borrow either or both concepts with a note of attribution.

Raymond McVay +William Black, I proposed something similar in terms of Command Modules myself back in the day. The article actually illustrates +Winchell Chung 's point rather well.

John Reiher Nice thing about a VIN Box is that the issuing authority can put a time limit on how long it's good for before you have to go and get it renewed. Think of it as license plates for spaceships.

Winchell Chung Wow, lots of good ideas here.

I had dismissed the idea of using the habitat module for ship ID because a robot unmanned ship would not have one. But +William Black idea of command module has some appeal. If you narrow it to the module that has actuators and computers controlling the various other modules. So the command module is a box with cables or BlueTooth connections to control all the other ship modules. Plus an I/O port for the captain to issue commands to the command module, and so the CMOD can report status reading to the captain.

This would indeed make the CMOD the sine qua non of a spacecraft, worthy to be indelibly embossed with a serial number usable as the ship's identity.

Plug in a human usable control panel into the CMOD I/O for human manned ships.
Plug in an AI interface connecting the AI computer to the CMOD I/O for AI manned ships.
Plug in a sequencer interface connecting a moron computer to the CMOD I/O allowing the moron computer to execute a pre-programmed set of commands as if it were a space-going player piano.
Plug in a radio interface connecting a radio to the CMOD I/O for a remote-controlled drone ship.
Or any combination of the above

+Raymond McVay's idea of the ID of the ship tied to its AI has merit. The only thing is AIs are absurdly easy to clone (once you crack the DRM copy protection). It is hard to stamp a serial number on software running in a computer. Especially if the AI software is self-modifying, as human being are. How does one distinguish one AI from another?

+John Reiher idea for a VIN Box is probably my favorite. It is very much the sort of thing that banking and insurance companies would come up with. And the requirement for renewability makes perfect sense. However it must always contain a legal ID, for liability purposes. Much like automobiles. If somebody crashes their car into a building or something else expensive, then flees the scene on foot, the car's license plate may be expired but it still allows the police and building owners to discover who is liable for the damages.

This also vaguely reminds me about ship transponders in the Traveller role playing game. They constantly broadcasts the ship's unique ID and location. Civilian starships are legally required to have transponders always turned on, unless there are extenuating situations. Such as pirate corsair ships in the area, using the transponder to home in on their prey.

William Black I was thinking about that very same problem with AI's last night and this AM, as Winchell points out AIs are absurdly easy to clone. It is hard to stamp a serial number on software running in a computer. How does one distinguish one AI from another?

All I could come up with is this: A ship's AI is hardware configured so its inputs and outputs must plug in through an interface that is part of the VIN box, and these are highly tamper resistant. Both the AI and the VIN box are hardened in various ways, with both physical and with hardware/software safeguards against tampering.

In other words a ship's AI cannot function without a registered spacecraft with legal ownership.

Under the heading of future crime: Cracking the AI-VIN box security feature.

It's probably not impossible to do, but it might be very, very, difficult to do — a man with the right skill set could sell his labor at a high price on the black market. Skills would be comparable to a high-level safe-cracking expert combined with high level hardware/software expertise.

I was thinking of similar safeguards in relation to nuclear pulse systems in my setting as well.

John Reiher Actually, the current lines of research into AIs indicate that they will be very hard to "clone" as they will be as much hardware as human minds are. You can copy the data, but not the mind. (Unlike human brains, which have no I/O ports.)

But I do like +William Black's idea of tying the AI to the ship's VIN Box. That way the ship's AI is really part of the ship. But that lasts until AI's get equal rights. Then they will operate independently of a ship's VIN Box. Of course if they do get equal rights, there's nothing stopping one from buying their own VIN Box and leasing their "ship" to whoever can afford their terms.

Alistair “Cerebrate” Young I'm pretty sure in my 'verse the legal identity of a starship is vested in the leatherbound data rod/smart-paper folio in a safe in the captain's office (or welded in a suitable location on a drone ship), which is to say, the certificate of registry.

Components may come and components may go — and the Flight Administrator will faithfully update said folio's documentation of said components[1] each time or be hauled up in front of an Admiralty Court for seriously violating the Imperial Navigation Act — but CMS Gorram Freebies, Hull Number Eleven-Oh-Seven-Four-Two-Niner remains CMS Gorram Freebies, Hull Number Eleven-Oh-Seven-Four-Two-Niner whatever gets replaced up to and including said hull as long as it's operating under the same non-decommissioned certificate of registry.

(Of course, the bank that holds the note on your starship may have its own documented ideas on what exactly it holds the note on, and should the registry and the mortgage get out of sync on this point, your life may become... interesting.)

(ed note: a visit from the Repo Man)

((Equally of course, the spec plat in the engineering computers will also have its own documented ideas about what the ship's made of and can be expected to do, and when it gets out of sync, your life may also become interesting.

Also, depending on how out of sync it is with reality, potentially hot, noisy, and short.))

(ed note: spacecraft undergoes rapid explosive disassembly because the engineering computer's mathenatical model of the spacecraft did not match reality)

[1] I like to imagine this as a nice hierarchically-organized document that begins with: Starship, free trader, Kalantha-class: 1
And ends with something like: Rivet, 4mm: 18,297

John Reiher Papers can be forged, data copied and manipulated. You need something that breaks if someone tries to open it. That's where the VIN Box comes in. Trying to crack one open is a operation in futility. The ID in the VIN Box is one half of a public encryption key that mates with a governmental key to validate your ship. If the two don't mate, you have a hacked box.

The Key that produces the public keys is private and in its own black box. The guys in the Ship Registry office just know that they have theses boxes that need ship's names, and that they already have IDs ready to go.

As an aside, you realize that we've been using modular space ships ever since we started building them. What do you think the Saturn V is? It's a one use modular rocket that you can put anything you want on top of the booster section and throw into orbit.

So that implies that a modular ship will have very good connections between the parts. On par with what a multi-stage rocket uses. Docking connections are just that, docking connections, and may or may not have the necessary structural strength for a space ship.

The type of connections that make up a modular ship most folks would call "permanent" but to folks in the business, they are temporary.

Alistair “Cerebrate” Young Side note on registries and transponders: of course, the Worlds being a not-exactly-unified group of polities, actual requirements on these points vary widely.

Even leaving aside the anarchic Rim Free Zone (the entirety of whose admiralty law could be summed up as "try not to hit anything that might complain"), the Accord on the Law of Free Space leaves it at the minimal "you should have a certificate of registry and a transponder that will squawk it out when queried". Local regulations, on the other hand...

(The Empire, for example, which holds that sovereignty begins with the individual, will happily accept self-signed registries and doesn't require much transponding, although lacking functionality in this area may leave you restricted to operating VFR and staying outside all regions of controlled space.

The other end of this particular bell curve is the Hope Hegemony, which wants transponders willing to disgorge pretty much any information you can think of on demand, including your code-signed visa from the Hegemony Bureau of Navigation and remote-slave ackles for your starship, and legally defines any vessel without such as "debris, subject to salvage and/or destruction at discretion".)

(ed note: "ackles" is Access-Control List. "Remote-slave ackles" means to grant the Hope Hegemony government the ability to seize command of your ship and fly it by remote control, at their whim)

John Reiher So +Alistair Young in this setting, if you want to be safe, you register your ship in the Hope Hegemony, which is the most restrictive polity and go elsewhere to do your business. (Or at least get a HH compatible VIN Box). 

Alistair “Cerebrate” Young Well, it depends on what you mean by "safe"...

Flying around most of the Worlds with an HH registry is kind of like sailing around Earth in a North Korean-flagged ship. It may make you welcome in Pyongyang, but the association with the local crazies may not do you any favors elsewhere...

(Common wisdom would probably suggest either a registry from one of the inoffensive, minor, single-system polities that hasn't really had the chance to offend anyone yet, or possibly an Imperial registry on the grounds that while they have offended lots of people, they're also notorious for sending gunboats after people who trouble them and theirs, so...

Neither of those is safe everywhere, though, which is why the Starfall Arc Free Merchant Confraternity suggests that the wise smuggler free trader goes to the trouble of discreetly procuring a number of suitable registries...)

+John Reiher I'll admit to being rather cynical where uncrackable devices are concerned, mostly because over my past IT career I've spent an awful lot of time watching notionally uncrackable software and hardware both be cracked, commonly within about a week of shipping...

So, the way I look at it is the standard software security way — never trust the client. If the papers and their cryptosignatures and the ship's "biometrics" and whatever other information you can gather match a recent update of the issuer's copy of the database, you can probably be sure that they are who they say they are, or at least who the issuer thinks they are. If you don't have that handy to verify against... well, it's time to take your best guess. 

Winchell Chung +John Reiher said: of course if they (AIs) do get equal rights, there's nothing stopping one from buying their own VIN box and leasing their "ship" to whoever can afford their terms.

Hmmmmm, interesting. I am reminded of the Spline aliens from Stephen Baxter's Xeelee novels. They are whale-like aliens who genetically engineered themselves to be exceptionally good spacecraft. They then rent themselves out to other races as spaceships for hire.

This also reminds me of the brain and brawn ships from Anne McCaffrey's The Ship Who Sang. The parents of severely deformed babies are given the option of having the baby engineered into becoming a shell person. They are encased in titanium shells and given a brain-computer link. Among other things they can be plugged into a starship, making a living ship. The process is expensive so the shell people come of age with heavy debts which they must work off in order to become free agents.

McCaffrey said: "I remember reading a story about a woman searching for her son's brain, it had been used for an autopilot on an ore ship and she wanted to find it and give it surcease. And I thought what if severely disabled people were given a chance to become starships? So that's how The Ship Who Sang was born."

John Reiher +Alistair Young it's a given that most hackers can crack the security of simple systems, but I'm talking about an ID that's at least 128 characters long, using the entirety of the UNICODE font character set, you have almost 39,000 glyphs. It will take the heat death of the universe to crack that code.

It would be easier to get a bootleg one that's already registered and in the system.

Also, I'm using a setting where while there are multiple governments, they do have treaties with each other and have agreed upon a common ship's registry system just to keep the confusion down and to prevent what you proposed: Smugglers with multiple transponders.

+Winchell Chung I'm also reminded of Eric the brain in the jar from Niven's Becalmed In Hell and The Coldest Place. He was the ship and could be put into any vessel as best as I can remember. 

Alistair “Cerebrate” Young +John Reiher Well, that depends on your methods. 128 UTF-16 characters — well, that's basically a 2048-bit key, and on average, sure, an RSS key of that length will take 6.4 quadrillion years to crack by brute force .

But we've cracked lots of them in reality by other methods: usually involving taking advantage of things like patterned data (even using only printing characters, rather than the entire code space, will halve the effective entropy of the key); flawed algorithm implementations; finding the government's or corporation's back-door they so often have left for themselves; or information leaks because there's always good old rubber-hose (torture) or brown-envelope (bribery) cryptanalysis.

(It also assumes no-one's invented quantum computing or found a constructive proof that P=NP.)

One advantage of checking against an external database is that, in theory, cracking the client won't do you any good because the database still won't match and cracking the db should be much harder without physical access, etc. It's just that that then begs the question of why it's worth bothering to secure the client in the first place.

...of course, verifying against an external db has its own issues of synchronization and light-lag, such as when Cap'n Harbatkin squawks an id that isn't in your local database and claims, on asking, that he updated his registry back on Flern and it's not his problem that you haven't got the updates yet. Is he a smuggler with a random number generator, or is he a legitimate trader whose lobby group will be screaming for your head on a spike if you hold him while you query Flern for verification and wait for an update to come back...?

John Reiher True, there is that, but still, you'd have to be a government or a corp to afford the computing power to crack one... OK, or have a botnet that doesn't go down because someone starts downloading pr0n.

But now we're falling into one-upmanship and that game never ends well. So I'll concede that there ain't no such thing as perfect security. Just good enough that only professionals and governments will bother to try to crack it.

Well, one solution is that your external db contains registry IDs that haven't been issued yet, but has a VIN Box waiting to be issued.

Of course that means if you can get one of those VIN Boxes illegally, then you got a valid ID. Or better yet, bribe the registrar of some backwater world to issue you three or four for use as you see fit.

From private conversation on Google Plus (2015)

(ed note: Veyndayk, Velmeran, Dveyella, and Keth are Starwolves. They are in the business of capturing warships of the stodgy bureaucratic interstellar Union and selling said warships back to the Union.)

     Soon they saw that it was Veyndayk, the cargo supervisor.
     "Business done," he said, stepping up to join Velmeran and Dveyella at the rail where they had been watching traffic pass on the level below.
     "Did you sell Keth back to the Sector Commander?" Velmeran asked.
     Veyndayk laughed. "No, although that might be a good use for old Starwolves. Farstell Freight and Trade bought back a shipment of clothing, conveniently packed in their own shipping containers. And (Union) fleet ordnance has just now payed us a finder's fee for an intact cutter."
     "A cutter?" Velmeran asked. Cutters were the smallest of the military ships, hardly bigger than a transport, and generally used only for police work.
     "My little joke," Veyndayk explained. "We took two intact cutters as riders on salvaged battleships, and one we have had sitting in a forward bay for the last year. We took them apart down to the smallest bolt and rebuilt the ships by taking parts at random. Now I am going to collect finder's fees on those ships in three different ports. That should give the boys in fleet ordnance fits, when they cross-check serial numbers of those parts."
     That appealed to Dveyella, who liked frustrating Union officials best of all. "You know, they will not be able to use those ships until they take them apart and rebuild them as they originally were."
     "You laugh, but that is probably the truth," the cargo officer said.

From THE STARWOLVES by Thorarinn Gunnarsson (1988)

Bare Bones Example

This is an interesting design example from the always worth reading Bootstrapping Space blog by Chris Wolfe. It is mostly centered around estimating a mission delta V and sizing a propulsion system to fit, but his thought processes are interesting.


All previous posts described all-chemical systems that could be built and operated profitably in the near term. This one focuses on electrical propulsion systems.
The defining features of most electric propulsion:
 - High efficiency (high Isp)
 - Low thrust
 - High power requirements
 - Long trip times
 - Long operating life

     I chose a specific paper (Frisbee, Mikellides) to examine since the authors thoughtfully included most of the interesting parameters for a reusable Nuclear Electric Propulsion (NEP) Mars cargo tug. I don't really dive into how to calculate this for yourself because the problem is quite difficult without modeling software.

     It all comes down to the details; the question of NEP vs. Solar Electric Propulsion (SEP) vs. Chemical depends on the specific mission goals and technologies used.

The summary:
23 tons dry mass for a nuclear-electric tug of ~6 MW thermal / 1.2 MW electric
64 tons cargo capacity from low Earth orbit to Phobos-Mars orbit
Just under 40 tons of water propellant for the outbound trip and another 7.2 tons acquired at Phobos for the return
2.2 years outbound, slightly less inbound
Two round trips between thruster refits, five round trips between reactor refits

(skip to the next section if you are already familiar with electric propulsion)

     The general idea is to use electrical power to dump energy into a propellant and then release it at very high speed.

     The simplest of these is Electrothermal.
     An electric current produces heat and the propellant is passed through it. First up is the resistojet, where a resistor somewhat like an incandescent lightbulb filament is heated and then the propellant is pumped past it. These are common devices in the RCS systems of satellites. Second is arcjet, which passes an electric arc directly through the propellant instead of through a resistor. These can reach higher Isp because they can heat the propellant beyond material limits for resistor elements.
     Efficiency is moderate (Isp of 500 to 1000, well above any realistic chemical system but easily an order of magnitude lower than the most efficient electrics). Preferred fuels are low atomic weight with no particulates (hydrogen, water, ammonia, hydrazine). Thermal power is most efficient when there are few degrees of freedom for the molecules so more of the energy can be applied to macroscopic motion, so hydrogen or hydrogen-rich propellants are ideal.

     Next up is Electrostatic.
     These are the 'traditional' ion thrusters, NSTAR, Hall effect, etc. The propellant is ionized by a strong electrostatic field (with some variations) and then the ions are accelerated by a negatively-charged grid or electrode. An electron gun is used to neutralize the electric charge of the ion beam and keep the spacecraft electrically neutral.
     Efficiency is high. with Isp ranging from 1000 to 5000 for most designs and a few reaching as high as 10,000. Preferred fuels are high atomic weight gases or elements with a very low first ionization energy; argon and xenon are frequently used. Sometimes iodine or metals like tin, magnesium or sodium are used, while caesium and rubidium can be used in a FEEP thruster.

     Lastly, Electromagnetic.
     These are the 'new generation' plasma thrusters like VASIMR, MPD, PIT, etc. Note the idea is not new, it's just that these technologies have been getting a lot of press lately. In fact, a pulsed plasma thruster (from this family) was the first electric thruster flown in space. In these devices the propellant is ionized by arc discharge, microwave heating or other means and accelerated by a magnetic field rather than an electrostatic grid.
 Efficiency is typically variable but can be very high, with an Isp of 1000 to 30,000 (most commonly about 1500 to 6000). Preferred fuels are the same as electrostatic thrusters for the most part, with lithium making an appearance in MPD thrusters.

     All three families have been in use for decades, while each family has relatively recent members pushing the limits. All share common fundamental physics with regard to their efficiencies. Sources of loss are in the power processing units, ionization energy of the propellant, dissociation energy of the propellant if it is a molecule rather than a pure element, ion impacts with the grid or body and in the charge density and geometry of the exhaust.


     Of key concern for a reusable vehicle is the propellant should be available in space. Xenon, argon and nitrogen are available in the atmosphere of Mars. Small amounts of nitrogen may be available in lunar cold traps or bound in soils of Ceres and many C-type asteroids. Water appears to be widely available. Alkali metals would be available on the Moon and in most asteroids.

     Another critical factor is that the thruster system should be reliable over the long term.  Electrostatic systems have already demonstrated very long operational lifespans in the range of 5 to 10 years of active thrust. Electromagnetic systems don't yet measure up in demonstrated lifespan, but that is mainly because most systems of this type are low Isp / high thrust RCS components designed for relatively short operating times rather than main drive units designed to last a decade. As a practical engineering concern, there are difficult challenges in improving lifespan of systems with exposed electrodes or grids. Designs without grids, whether electrostatic or electromagnetic, have the potential for decades-long operation.

     A practical system will be able to service a cargo route in a reasonable time. To illustrate this, let's dive into a paper on a pulsed inductive thruster proposal that includes payload transits to Mars, Saturn and solar escape. A reusable cargo tug with about 64 tons of payload and a 2.2-year Mars transit would require about 2 megawatts of electricity and about 275 tons fueled mass at departure. Propellant would be plain water, though ammonia works as well. If propellant supplies are available at Mars then we require only 1.2 megawatts and 165 tons fueled mass. Of that, the tug itself is just under 23 tons. Enough spare components would be included to make two round trips between thruster refits. A refit would mean swapping out the entire thruster pod with a fresh assembly, quick and easy.
     2.2 years is a long trip, but it is right at the synodic period for Earth and Mars. A cargo tug launched at one opportunity would arrive just in time to go / no go the crew launch at the next opportunity. Part of the problem is that a low-thrust vehicle has to produce about 16 km/s of dV for this trip, far more than a high-thrust chemical system requires thanks to the Oberth effect. Assuming an Isp of 6000 seconds the fuel mass ratio is 23.81% or 39.285 tons of water. An empty return trip would require 7.2 tons of propellant. Put another way, each ton of fuel delivers 1.6 tons of payload. Contrast that with my chemical tug's ratio of 0.8 tons of payload per ton of fuel and you can see the advantage; twice the mass delivered for the same quantity of fuel. Of course, the two approaches trade off costs between dry structure and fuel; electric propulsion is not automatically better but it certainly lets you do a lot more with the same starting mass.

     Power can come from two sources, solar or nuclear. The tug described above is nuclear; its reactor should be good for five or possibly six round-trips out of the box but could be designed for a much longer lifespan of 40 to 50 years with a bit more mass. This is partly because the same propulsion system is intended for exploration missions to the outer planets, where solar power is minimal.

     A Mars cargo tug could certainly use solar power instead, with a lifespan in the 20 to 30 year range and less complicated refitting / disposal. The ~19 tons of reactor, radiators and conversion hardware would be replaced by very large solar panel arrays of about 30 tons, 3.0 MW at Earth beginning-of-life yielding 1.2 MW at Mars end-of-life, with a whole-system specific power of 100 W/kg, 20-year useful life, 20% degradation and 50% of Earth-normal power available at Mars. If Photovoltaics (PV) refits are available every two or three trips then the allowance for degradation can be reduced to perhaps 10%, saving about 3.5 tons.

     Reactors also degrade over time as their fuel decays; some designs such as traveling wave, drum reflector and pebble bed can level the power output over time by only burning part of the nuclear fuel load at any one time. These designs can be life-extended by including more fuel during construction and / or by replacing fuel elements. Since the nuclear fuel is only a tiny fraction of a reactor's mass, this life extension adds very little mass to the overall system. This is the same reason why increasing a reactor's power takes less mass than increasing a solar panel array's power; for low power outputs the solar panels are nearly always lighter due to the reactor's heavy power conversion equipment and shielding, but as the power output grows the reactor eventually beats PV. As you can see from the above example, 1.2 megawatts at Mars after 20 years is firmly in nuclear territory given current state of the art solar performance. Even so, I would bet there is room in the design space for solar PV to be competitive at this power level, design life and solar distance.

Future work:

     I think the next step is to work up some EML2 to Phobos tether-capture cargo runs and see how they compare to the baseline LEO to LMO mission. Keeping used nuclear reactors out of LEO is a good idea, as is keeping large solar arrays out of the Van Allen belts. I'll try for an electric interplanetary tug with similar payload to my chemical tug. These would have the added bonus of providing abundant power while parked; there may be a case for a set of tugs such that one is always at Phobos providing megawatt-scale electrical power.


     I need to continue on the topic of electric propulsion. The previous post was a lot of words but not a lot of meat. I felt it was too weak to stand alone, particularly as a part of this series where I am trying to focus on a realistic near-term plan for cargo transport. If you are interested in more background information I'd start with the Wikipedia page on electric propulsion and follow up with a look at the Atomic Rockets engine page. Another good look in the context of interplanetary travel is this paper (Hellin), while a deep look at relevant equations can be had in this paper (Keaton).

     One interesting result is a general rule to find required thrust given average acceleration. Google failed me on finding an exact solution, but it looks like there is a simple approach that is within 1% of the target value.

     I eventually settled on a design massing 33.4 tons, 1.6 MW solar-electric, Isp 6,000 and 40 N thrust using PIT thrusters with water propellant.

     Let's look at an electric tug with payload comparable to my reference tug, both a solar PV and a nuclear version. The main routes for this vehicle will be between LEO, GEO, EML1/2 and Mars orbit. Unlike the chemical tug we can't get much out of the Oberth effect, so the delta-V requirements are higher. Just like the chemical tug, the LEO to EML1 leg has the highest dV requirements (about 7km/s), so if we design for that case then the other trips will be faster, carry more cargo or burn less propellant.

     A key design factor here is trip time. If we throw enough power at the problem we can get to EML1 in the same amount of time as a chemical rocket, but that is a poor use of the mass. We need to decide how long we are willing to wait for the cargo and design enough thrust into the ship to make the trip in that span. I'm going to suggest four weeks to EML1 as a reasonable compromise, so let's see the consequences of that choice.

Estimating thrust requirements for average acceleration

     To apply 7km/s of delta-V in 28 days we need to make an average acceleration of 2.89 mm/s. To allow some wiggle room let's assume we can only thrust 90% of the time, meaning now we need 3.22 mm/s. Since this is our average acceleration, we need to find either the initial or final acceleration to find the thrust of the propulsion system. To do that we will first need to know the vehicle's propellant mass fraction, so let's take a few test cases at Isp of 3000, 6000 and 10,000.

     Propellant mass fraction (Mf) is equal to 1 - e ^ (- dV / Ve), where dV in this case is 7000 m/s and Ve is Isp * g. See the rocket equation page for more details.
Isp 3000 -> Mf of 0.21175
Isp 6000 -> Mf of 0.11217
Isp 10000 -> Mf of 0.06890

     It would be nice if the average acceleration also matches up with the midpoint of fuel consumption, but somehow I doubt it. Let's find out.

     Given a dry mass of, say, 10 tons and an Isp of 3000, the fueled mass is ( 1 / ( 1 - Mf ) ) * dry mass, or 12.686 tons. When half of the fuel is burned the craft masses 11.343 tons. The target acceleration is 0.00322 m/s, so the required thrust is 36.52 newtons. Thrust is mass-flow (mdot) times exhaust velocity, so mdot is 1.2414 grams per second. That rate of propellant consumption would require 25 days to empty the tank, or 27.825 days after accounting for our 90% duty cycle.

     That surprises me. It's not exact but it is close enough for exploratory work. The case of 10,000 Isp works out to 27.947 days, so it looks like this general rule is valid across a fair range of Isp values. I also spot-checked some different mission dV values and found similar agreement, always within 1%. If anyone out there knows of an exact solution I would love to hear it.

     To calculate this yourself you need your mission dV, Isp, thrust duration and a test mass. If you set the test mass to 1kg (or 1t) then you can find a multiplier to use for different dry masses. The relationships are linear.

The required acceleration a will be dV in meters per second divided by thrust duration in seconds.
First, find fuel mass fraction, which is 1 - e ^ (- dV / Ve).
Convert to dry mass fraction Md, which is -Mf + 1
Convert to the 'gear ratio', which is simply 1 / Md
Multiply by dry mass M1 to get fueled mass M0 and note this value.
Find the fuel mass by taking M0 - M1 and note this value.
Find the 'halfway point', which is half the fuel mass plus M1; let's call this Mh.
Find the thrust F, which is Mh * a. This is the value you are looking for.
Find mdot, which is thrust divided by exhaust velocity, or F / (g * Isp ).
Find the real thrust duration, which is fuel mass divided by mdot. This should be within 1% of your stated thrust duration; if it is then the average acceleration value is accurate enough to use.

     If you have a known spacecraft (known dry mass and fuel mass, known thrust), you can use thrust divided by (dry mass plus half the fuel mass).

Electric tug design

     To align with the chemical tug, let's target a payload of 40 tons from LEO to EML1. Note that EML2 is a better target, but for purposes of comparison I'm using the LEO to EML1 trip as the most costly trip in the set. As mentioned above, we need to deliver in 28 days or provide an average acceleration of 3.22 mm/s. I don't have an exact solution, so I can't solve the problem in a single step. That's fine; spacecraft design is an iterative process.

     Let's assume an electric thruster at Isp = 6000 and mission dV of 7000 m/s. Also assume a one-way trip (meaning fuel is available at both endpoints). Power alpha is assumed to be 18 kg/kW, whether that be nuclear or long-life solar. Thrusters will be the NuPIT design shown in the last post, using the design values for the 5 N, 200 kW unit at 2.75 kg/kW (550 kg per thruster).
     As a first guess let's try eight thrusters, 1.6 MW. That's 33.2 tons, for a dry mass without tanks of 73.2 tons. We will need approximately 10 tons of liquid water propellant; using a tankage fraction of 2% would be reasonable in this case, so tack on 200kg for tanks for a total dry mass of 73.4 tons. Actual propellant load is 9,273 kg, so tankage is sufficient. The half-fueled mass is 78,037 kg and approximate average acceleration is 0.513 mm/s. We're not even close. Trip time would be 157.9 days, or 2.3 one-way trips per year.
     Maybe 20 thusters / 4 MW? Power alpha would improve to about 15, yielding 71 tons of power and propulsion. 40 tons of payload and perhaps 0.4 tons of tankage gives a dry mass of 111.4 tons, fuel mass of 14.1 t and average acceleration of 0.844 mm/s. This is clearly not going our way. Trip time would be 106.3 days, or 3.4 one-way trips per year.
     Let's aim much higher, 50 thrusters / 10 MW. Power alpha would continue to improve to about 14, yielding 167.5 t of power and propulsion. 40 tons of payload and 0.6 tons of tankage gives a dry mass of 208.1 tons, 26.29 tons of fuel and an average acceleration of 1.13 mm/s. Trip time would be 71.7 days or 5.1 one-way trips per year.

     Clearly, short trip times require increasingly absurd power levels. Matching the payload size of a chemical thruster with the 1.6 MW version means only making one round-trip per year. In fact, looking at that version of the ship, if we eliminate the payload entirely the highest acceleration the ship can make is 1.2 mm/s on its last gasp of propellant. Since fuel mass, dry mass, power and thrust are all linear relationships* that means no matter how we scale up the ship it can never get better than this. (The power system alpha does actually get better as we scale up, but moving a ship that masses several times your payload is inefficient and extremely expensive.)

     One thing we can do is increase the thrust of each propulsion unit, which usually means decreasing the Isp significantly. Let's look at a VASIMR thruster for comparison, since I have some data on performance at different Isp levels handy. A VASIMR thruster at 200 kW and 6000 Isp produces about 4.75 N of thrust, a fairly close match to the NuPIT. We need about six times that thrust (28.5 N), which occurs right at an Isp of 1000. That would bring the 8-thruster 1.6 MW vessel up to about 230 N of thrust. However, dropping the Isp so dramatically brings the fuel fraction just over 50%. That pushes our dry mass up to 75t, fuel mass to 78.1t and nets us only 2.0 mm/s average acceleration. It's a 40.5 day trip or 9 trips per year, but now we are burning more fuel than the chemical tug thanks to our drastically higher dry mass. Still no net benefit to be had.

Putting the tug to work

     Let's look at what an electric tug actually saves: propellant. In a fully functional ecosystem of cis-lunar services propellant is fairly plentiful. The speed, convenience and throughput of chemical vehicles far outweighs the efficiency of ion vehicles in this environment. Where an electric tug shines is in the buildup phase, where all of the propellant is coming from Earth. The tug would save money during a critical part of the project. What that means is we do not need to survive dozens of Van Allen belt transits over two decades, we just need to make a reasonable number of trips over two or three years. We also don't need to standardize on the same payload sizes as the chemical tug, nor do we need to make trips in 1 month. I would say that using the same power system alpha for the solar version as I do for the nuclear version is very pessimistic; these vessels would not need to function at Mars orbit, though they do need significantly thicker front-glass shielding on the panels than other craft.

     So, a lunar ISRU plan would still start with a single chemical tug / lander as described in part 1. Using performance for the detailed reference tug, a 15-ton package can be delivered from LEO direct to the lunar surface. This will be 12.4 tons of ISRU equipment and 2.6 tons of spares (2.1 year supply). Refilling tug 1 will take 6 months, after which it can deliver 33 tons to EML1.
     In the meantime, an electric tug (call it tug A) will deliver a 9-ton fuel depot (135 ton capacity) to EML1. Let's use our 40 N / 1.6 MW / 33.4t / 6000 Isp vehicle from above. It does the job in about 92 days, which means there is a window of three months after the launch of the first ISRU package to get the 33.4t tug, 5.36t propellant and 9t payload into LEO.
     At the first lunar launch, 33 tons are delivered to the EML1 depot. Tug A will collect this and head to LEO, taking 127 days and consuming 7.45t of water. During this trip a LEO depot is launched, identical to the one at EML1. The tug turns back around and heads for EML1, taking 4.22t of water for the return trip and leaving 21.33t of cargo in LEO. This could be a mix of surface samples and water as desired. Let's assume five tons are samples and the rest is fuel.
     The return trip takes 72 days, during which tug 1 will have delivered another 33 tons to the EML1 depot. Tug A repeats its performance, returning to LEO with a full load of 23.78t water and another 5t of samples. At this point we are at 598 days elapsed since start of ISRU operations, which should be enough time to settle on and construct additional hardware to expand the lunar surface capacity.
     This is significantly longer than the all-chemical scenario and has an IMLEO of 129.71 tons, within a few tons of all-chemical. Hardware costs are higher since more of the mass is spacecraft and much less of it is fuel. The main benefit is that schedule pressures are greatly reduced; final design, construction and testing of the second round of ISRU plant is allowed more than a year and a half of time rather than two months. More operational data is available and the tolerance for mistakes or inefficiencies is higher. Another benefit is that this profile includes depots in LEO and at EML1; even if things do not progress beyond the first ISRU package the infrastructure is still useful for this and future projects.
     This baseline hardware could continue to deliver 21 tons of cargo to LEO every ~200 days for about a decade, eventually reaching 426 tons over 20 trips at a cost of 141 tons of Earth mass or a leverage of about 3 to 1. Things improve if we continue to expand, since about 44% of that mass was fuel to get the first ISRU plant in position; additional ISRU hardware is delivered using lunar propellant.

     The next phase would be to send more ISRU hardware. Tug A can pick up 33 tons at EML1, deliver 19.18t of net payload to LEO over 127 days, pick up a 17-ton package and head to EML1 in 109 days. All of the required propellant is lunar and picked up at EML1. Round trip time is 236 days (a bit under 8 months). The harvesting process run by tug 1 has a shorter turnover time of 6 months, so on average an extra 19 tons is accumulated at the depot. That's not quite enough to provide for a cargo landing, so tug A may not always be bringing a full load of cargo to LEO (meaning shorter round trips in practice).
     An alternative might be to use 12-ton packages that will fit into a Falcon 9 for cheaper launch costs; the delivery time for that is 98 days. If less cargo is returned to LEO then that trip time can be shortened as well; for example, 6 tons of return cargo plus round-trip fuel would make each leg of the trip take 98 days, or 196 days round-trip. Each 6.5-month trip would deliver another 10 tons of ISRU with two years of spares. Two electric tugs could deliver 80 tons of ISRU capacity in 26 months, roughly a single Mars synodic period. That would place 95 tons of ISRU with expected output of 950 tons of propellant annually at a cost of 9.5 tons of spares. Net propellant delivered to EML1 would be 505 tons annually, or could be 168 tons to LEO annually with chemical tugs. The annual demand for spares (both ISRU and depots) can be met in a single hardware run with minimal fuel costs, leaving 3/4 of the electric tug schedule open for assignments like delivering new chemical tugs or GEO debris retrieval (a mission that avoids the majority of the radiation belts and prolongs the tug's useful life).
     The total phase 2 IMLEO would be 134.4 tons, all of it hardware. Lunar mass to LEO during this period would only be another 48 tons since capacity is focused on buildup. This phase would run for 26 months, or a total of 46 months since first launch.
     Ongoing maintenance would require approximately 12 tons per year. The initial depots would be insufficient, so we need another 27t of hardware for fuel storage. If we rate all flight and depot hardware with a 10-year lifespan and pro-rate the replacement mass then we need an additional 12.2 tons annually (24.2t total). Depending on how the output is allocated, this could be considered an ongoing leverage of 23.5 tons in EML1 per ton IMLEO or 7.8 tons in LEO per ton IMLEO. Another way to look at it would be as a fuel supply for three manned Mars missions covering four synodic periods (104 months), or a full ISRU program length of 150 months (12.5 years). Overall Earth mass to LEO is then 511.15t to harvest 5,501.7 tons of gross lunar propellant, yielding 2,924.6 tons net lunar propellant at EML1. That's a gear ratio of about 5.7 to 1. If you are only interested in delivering fuel to LEO then you can net 972.9 tons, still a favorable 1.9 to 1 mass ratio. 511 tons is a lot of mass to launch, but only three payloads require a heavy lift vehicle: the initial chemical tug stack (62t fuel and 22t hardware, split across two Falcon H) and the two electric tugs (33.4t each, also requiring a Falcon H unless they can be built in parts and flown on two Vulcan launches). The remaining 360 tons would be delivered by 30 Falcon 9 launches, or by some combination of any price-competitive launchers with at least 12 tons of payload.

     Launch costs would be roughly $2.1 billion. Hardware would run another $6.7 billion (at $15m per ton). Operations might cost $125-$250 million. Call it a total of $9 billion over about 15 years (12.5 years of operation plus 2.5 years of r&d, manufacturing and testing). Overall cost of fuel at EML1 would be $3,094.44 per kg, about $3.1 million per ton and expected to decline to $0.9 million per ton in the long run. Savings are about the same as the all-chemical approach, a bit over $4.5 billion vs. NASA baseline. Additional savings could be realized by using the chemical tugs as cargo haulers to and from Mars as described in part 2, resulting in excess capacity that could be sold or used for other purposes. One of those purposes might be ISS reboost and water supply for life support. Another might be developing a significant water supply on the Moon for growing food, in support of manned missions.


When a spacecraft built for humans ventures into deep space, it requires an array of features to keep it and a crew inside safe. Both distance and duration demand that spacecraft must have systems that can reliably operate far from home, be capable of keeping astronauts alive in case of emergencies and still be light enough that a rocket can launch it.

Missions near the Moon will start when NASA’s Orion spacecraft leaves Earth atop the world’s most powerful rocket, NASA’s Space Launch System. After launch from the agency’s Kennedy Space Center in Florida, Orion will travel beyond the Moon to a distance more than 1,000 times farther than where the International Space Station flies in low-Earth orbit, and farther than any spacecraft built for humans has ever ventured. To accomplish this feat, Orion has built-in technologies that enable the crew and spacecraft to explore far into the solar system.

Systems to Live and Breathe

As humans travel farther from Earth for longer missions, the systems that keep them alive must be highly reliable while taking up minimal mass and volume. Orion will be equipped with advanced environmental control and life support systems designed for the demands of a deep space mission. A high-tech system already being tested aboard the space station will remove carbon dioxide (CO2) and humidity from inside Orion. Removal of CO2 and humidity is important to ensure air remains safe for the crew breathing. And water condensation on the vehicle hardware is controlled to prevent water intrusion into sensitive equipment or corrosion on the primary pressure structure.

The system also saves volume inside the spacecraft. Without such technology, Orion would have to carry many chemical canisters that would otherwise take up the space of 127 basketballs (or 32 cubic feet) inside the spacecraft—about 10 percent of crew livable area. Orion will also have a new compact toilet, smaller than the one on the space station. Long duration missions far from Earth drive engineers to design compact systems not only to maximize available space for crew comfort, but also to accommodate the volume needed to carry consumables like enough food and water for the entirety of a mission lasting days or weeks.

Highly reliable systems are critically important when distant crew will not have the benefit of frequent resupply shipments to bring spare parts from Earth, like those to the space station. Even small systems have to function reliably to support life in space, from a working toilet to an automated fire suppression system or exercise equipment that helps astronauts stay in shape to counteract the zero-gravity environment in space that can cause muscle and bone atrophy. Distance from home also demands that Orion have spacesuits capable of keeping astronaut alive for six days in the event of cabin depressurization to support a long trip home.

Proper Propulsion

The farther into space a vehicle ventures, the more capable its propulsion systems need to be to maintain its course on the journey with precision and ensure its crew can get home.

Orion has a highly capable service module that serves as the powerhouse for the spacecraft, providing propulsion capabilities that enable Orion to go around the Moon and back on its exploration missions. The service module has 33 engines of various sizes. The main engine will provide major in-space maneuvering capabilities throughout the mission, including inserting Orion into lunar orbit and also firing powerfully enough to get out of the Moon’s orbit to return to Earth. The other 32 engines are used to steer and control Orion on orbit.

In part due to its propulsion capabilities, including tanks that can hold nearly 2,000 gallons of propellant and a back up for the main engine in the event of a failure, Orion’s service module is equipped to handle the rigors of travel for missions that are both far and long, and has the ability to bring the crew home in a variety of emergency situations.

The Ability to Hold Off the Heat

Going to the Moon is no easy task, and it’s only half the journey. The farther a spacecraft travels in space, the more heat it will generate as it returns to Earth. Getting back safely requires technologies that can help a spacecraft endure speeds 30 times the speed of sound and heat twice as hot as molten lava or half as hot as the sun.

When Orion returns from the Moon, it will be traveling nearly 25,000 mph, a speed that could cover the distance from Los Angeles to New York City in six minutes. Its advanced heat shield, made with a material called AVCOAT, is designed to wear away as it heats up. Orion’s heat shield is the largest of its kind ever built and will help the spacecraft withstand temperatures around 5,000 degrees Fahrenheit during reentry though Earth’s atmosphere.

Before reentry, Orion also will endure a 700-degree temperature range from about minus 150 to 550 degrees Fahrenheit. Orion’s highly capable thermal protection system, paired with thermal controls, will protect Orion during periods of direct sunlight and pitch black darkness while its crews will comfortably enjoy a safe and stable interior temperature of about 77 degrees Fahrenheit.

Radiation Protection

As a spacecraft travels on missions beyond the protection of Earth’s magnetic field, it will be exposed to a harsher radiation environment than in low-Earth orbit with greater amounts of radiation from charged particles and solar storms that can cause disruptions to critical computers, avionics and other equipment. Humans exposed to large amounts of radiation can experience both acute and chronic health problems ranging from near-term radiation sickness to the potential of developing cancer in the long-term.

Orion was designed from the start with built in system-level features to ensure reliability of essential elements of the spacecraft during potential radiation events. For example, Orion is equipped with four identical computers that each are self-checking, plus an entirely different backup computer, to ensure Orion can still send commands in the event of a disruption. Engineers have tested parts and systems to a high standard to ensure that all critical systems remain operable even under extreme circumstances.

Orion also has a makeshift storm shelter below the main deck of the crew module. In the event of a solar radiation event, NASA has developed plans for crew on board to create a temporary shelter inside using materials on board. A variety of radiation sensors will also be on the spacecraft to help scientists better understand the radiation environment far away from Earth. One investigation called AstroRad, will fly on Exploration Mission-1 and test an experimental vest that has the potential to help shield vital organs and decrease exposure from solar particle events.

Constant Communication and Navigation

Spacecraft venturing far from home go beyond the Global Positioning System (GPS) in space and above communication satellites in Earth orbit. To talk with mission control in Houston, Orion’s Orion will use all three of NASA’s space communications networks. As it rises from the launch pad and into cislunar space, Orion will switch from the Near Earth Network to the Space Network, made possible by the Tracking and Data Relay Satellites, and finally to the Deep Space Network that provides communications for some of NASA’s most distant spacecraft.

Orion is also equipped with backup communication and navigation systems to help the spacecraft stay in contact with the ground and orient itself if it’s primary systems fail. The backup navigation system, a relatively new technology called optical navigation, uses a camera to take pictures of the Earth, Moon and stars and autonomously triangulate Orion’s position from the photos. Its backup emergency communications system doesn’t use the primary system or antennae for high-rate data transfer.

Spacecraft Parameters

For an given type of automobile, there are parameters that tell you what kind of performance you can expect. Things like miles per gallon, acceleration, weight, and so on.

Spacecraft have parameters too, it is just that they are odd measures that you have not encountered before. I am going to list the more important ones here, but they will be fully explained on other pages. Refer back to this list if you run across an unfamiliar term.

Habitat Module
The pressurized part of the spaceraft where people live. Included in Payload Section. Remember that Rockets Are Not Hotels. Unlike the Starship Enterprise a real spacecraft is a huge expanse of airless machinery with a tiny pressurized habitat module tucked away in a corner where people can walk around without spacesuits.
The part of the spacecraft that is its reason for existance. For a satellite booster, the payload is the satellite it is lifting into orbit. For a transport ship: habitat module, passengers, ship controls. For a warship: habitat module, crew, weapons, defenses, ship controls. For a robot freighter: robot controls and cargo. Some payload like cargo and crew are removable from the spacecraft. Some payload like weapons and habitat modules are fixed parts of the spacecraft. Included in Payload Section.
Engine or Thruster
The rocket engine that moves the spacecraft, and the empty propellant tanks. Included in Propulsion Bus.
Power Plant
Part that generates electricity. Included in Propulsion Bus.
Struture is the skeleton and skin of the spacecraft. Included in both Propulsion Bus and Payload Section.
Propellant and Fuel
Propellant or Reaction mass (remass) is what the thruster fires out the exhaust nozzle to create thrust. Fuel is the source of energy used to propel the propellant. Remember that Fuel Is Not Propellant. In chemical rockets, the chemicals are both propellant and fuel. In nuclear rockets the liquid hydrogen is the propellant and the uranium is the fuel. Included in Propulsion Bus.

Payload Mass (Mpl)
Mass of all the payload. For NASA vessels this is typically 26.7% of Dry Mass.
Payload Fraction (λ)
Payload mass as percentage of wet mass. Mpl / M
Structural Mass (Mst)
Mass of all the struture. For NASA vessels this is typically 21.7% of Dry Mass.
Propellant Mass (Mpt)
The mass of all the propellant in the spacecraft's propellant tanks. Does not include fuel that is retained after it is burnt, e.g., uranium fissioned inside a solid core reactor. For some calculations, you will use instead the mass of propellant that will be expended in a given maneuver.
Power Plant Mass (Mpp)
The mass of the electrical generation system. Includes any heat radiators. For NASA vessels this is typically 28% + 3.4% of Dry Mass
Thruster System Mass (Mts)
The mass of the rocket engines, including the empty propellant/fuel tanks. For NASA vessels this is typically 3.7% of Dry Mass
Propulsion System Mass (Mps)
Thruster System Mass + Power Plant Mass.
Inert Mass (Mi)
Mass of spacecraft with no propellant and no payload. Propulsion System Mass + Structural Mass.
Inert Mass Fraction (δ)
Inert mass as percentage of wet mass. Mi / M
Dry (Empty, Burnout) Mass (Me)
Mass of spacecraft with no propellant but with payload. Propulsion System Mass + Structural Mass + Payload Mass.
Wet (Total, Ignition) Mass (M)
Total mass of spacecraft. Propellant Mass + Propulsion System Mass + Structural Mass + Payload Mass.
Mass Ratio (R)
Ratio of wet mass to dry mass. Wet Mass / Dry Mass.
Parametric Mass Ratio (r)
λ + δ.
Propellant Fraction (Pf, PMF, or ζ)
Percentage of wet mass that is propellant. 1 - ( 1 / MassRatio )

Propellant Mass Flow (mDot or )

How quickly does the Thruster System drain the propellant tanks? Rated in kilograms per second.

mDot constrains the amount of thrust the propulsion system can produce. Changing the propellant mass flow is a way to make a spacecraft engine shift gears.

Exhaust-Velocity (Ve)

How fast does the propellant shoot out the exhaust nozzle of the Thruster System? Rated in meters per second. Exhaust velocity (and delta V) is of primary importance for space travel. For liftoff, landing, and dodging hostile weapons fire, thrust is more important.

Broadly exhaust velocity is a measure of the spacecraft's "fuel" efficiency (actually propellant efficiency). The higher the Ve, the better the "fuel economy".

Generally if a propulsion system has a high Ve it has a low thrust and vice versa. The only systems where both are high are torch drives. Some spacecraft engines can shift gears by trading exhaust velocity for thrust.

For a more in-depth look at exhaust velocity look here

Specific Impulse (Isp)
Another way of stating exhaust velocity. Exhaust Velocity / 9.81 where 9.81 = acceleration due to gravity on Terra in meters per second. Specific Impulse is rated in seconds. It is also a broad measure of the spacecraft's "fuel" efficiency.
Delta V or Δv

Spacecraft's total change in velocity capability. This determines which missions the spacecraft can perform. Arguably this is the most important of all the spacecraft parameters. Rated in meters per second.

This can be thought of as how much "fuel" is in the tanks of the spacecraft (though it is actually a bit more complicated than that).

Velocity Ratio
Δv / Ve
Thrust (F)

Thrust produced by Thruster System. Rated in Newtons. Thrust is constrained by Propellant Mass Flow. Thrust (and acceleration) is of primary importance in liftoff, landing, and dodging hostile weapons fire. For space travel exhaust velocity (and delta V) is more important.

Generally if a propulsion system has a high Ve it has a low thrust and vice versa. The only systems where both are high are torch drives. Some spacecraft engines can shift gears by trading exhaust velocity for thrust.

Acceleration (A)

Spacecraft's current acceleration. Current total mass / Thrust. Rated in meters per second per second. Divide by 9.81 to get g's of acceleration.

In space, a spacecraft with higher acceleration will generally not travel to a destination any faster than a low acceleration ship. But a high acceleration ship will have wider launch windows for a given trajectory.

Note that as propellant is expended, current total mass goes down and acceleration goes up. If you want a constant level of acceleration you have to constantly throttle back the thrust.

5 milligee (0.05 m/s2) : General rule practical minimum for ion drive, laser sail or other low thrust / long duration drive. Otherwise the poor spacecraft will take years to change orbits. Unfortunately pure solar sails are lucky to do 3 milligees.

0.6 gee (5.88 m/s2) : General rule average for high thrust / short duration drive. Useful for Hohmann transfer orbits, or crossing the Van Allen radiation belts before they fry the astronauts.

3.0 gee (29.43 m/s2) : General rule minimum to lift off from Terra's surface into LEO.

For a more in-depth look at minimum accelerations look here.

Thrust Power (Fp)
Power produced by Thruster System. ( Thrust × Exhaust Velocity ) / 2. Rated in watts.
Specific Power (Fsp)
Power density of spacecraft. Thrust Power / Dry Mass. Rated in watts per kilograms.
Specific Mass
Alpha of Thruster System. Thruster System Mass / Thrust Power. Rated in kilograms per watt.


Typically the percentage of spacecraft dry mass that is structure is 21.7% for NASA vessels.

What is the structure of the ship going to be composed of? The strongest yet least massive of elements. This means Titanium, Magnesium, Aluminum, and those fancy composite materials. And all the interior girders are going to have a series of circular holes in them to reduce mass (the technical term is "lightening holes").

Spacecraft Spine

Many (but not all) spacecraft designs have the propulsion system at the "bottom", exerting thrust into a strong structural member called the ship's spine. The other components of the spacecraft are attached to the spine. The spine is also called a keel or a thrust frame. In all spacecraft the thrust frame is the network of girders on top of the engines that the thrust is applied to. But only in some spacecraft is the thrust frame elongated into a spine, in others the ship components are attached to a shell, generally cylindrical.

If you leave out the spine or thrust frame, engine ignition will send the propulsion system careening through the core of the ship, gutting it. Spacecraft engineers treat tiny cracks in the thrust frame with deep concern.


However, I'm debating if the structures you cite as "keels" make sense when cross-referenced with "thrust frame".

For instance, ISS' truss isn't really a thrust frame—the station is very rarely under thrust, and when it is, it's usually from spacecraft or its own thrusters on the end of the Russian segment, which would actually make the whole main line of modules (Zarya, Zvezda, Pressurized mating adapter-1, Unity, Destiny, Harmony) the main "keel". The job of the truss in such a case is just to stop itself from flexing and hold the solar "wings" in place.

Similarly, there's other space vehicles which lack such a "keel" entirely, such as the DTAL concept or the Altair ascent stage design. In both cases, an engine is basically mounted to a pressure vessel (a prop tank for DTAL and a crew cabin for Altair) and then the rest of the structure "hangs" off of that pressure hull.

I might also note that these kinds of {keel-less spacecraft} will, in a rocketpunk setting, likely be confined to special-purpose craft—landers or scooters, spaceplanes, dedicated fuel tankers, and such. Most "typical" ships will probably have a bit more of spiney spaceframe.

The distinction I might make is "primary structure" and "thrust structure". The thrust structure is just the structural system to distribute the force of the engine, such as the F9 Octaweb. On the other hand, the primary structure is anything that serves a major structural role in the ship, analyzed as a system. In DTAL, it'd be the engine, the thrust structure that mounts that engine to the tanks, the tanks, and then the landing gear and such. For ISS, it's the outer hulls of the core "line" of modules, plus the truss. This primary structure might also be called an airframe or spaceframe.

Engineer Rob Davidoff (2014)

OK, forget what I just said. On top of the engine will be the thrust frame or thrust structure. On top will be the primary structure or spaceframe. The thrust frame transmits the thrust into the spaceframe, and prevents the propulsion system careening through the core of the ship.

The spaceframe can be:

  • A long spine/keel with the propellant tanks and payload section bits attached in various places.
  • A large pressurized vessel, either propellant tank or habitat module. Other propellant tanks and payload section bits are attached to main tank or perched on top.
  • Something else.

The engineers are using a pressurized tank in lieu of a spine in a desperate attempt to reduce the spacecraft's mass. But this can be risky if you use the propellant tank. The original 1957 Convair Atlas rocket used "balloon tanks" for the propellant instead of conventional isogrid tanks. This means that the structural rigidity comes from the pressurization of the propellant. This also means if the pressure is lost in the tank the entire rocket collapses under its own weight. Blasted thing needed 35 kPa of nitrogen even when the rocket was not fueled.

As Rob Davidoff points out, keel-less ship designs using a pressurized tank for a spine is more for marginal ships that cannot afford any excess mass whatsoever. Such as ships that have to lift off and land in delta-V gobbling planetary gravity wells while using one-lung propulsion systems (*cough* chemical rockets *cough*).

This classification means that parts of the propulsion bus and payload section are intertwined with each other, but nobody said rocket science was going to be easy.

In von Braun Round the Moon Ship the thrust frame (dark blue) is right on top of the rocket motors. The spaceframe (light blue) is a cage attached to the thrust frame. The rocket motors push upwards on the thrust frame, which pushes upwards on the spaceframe. The personnel sphere, hydrazine tank, and nitric acid tank are all basically inflated balloons hung on the spaceframe.

Getting back to the spine. Remember that every gram counts. Spacecraft designers want a spine that is the strongest yet lowest mass structural member possible. The genius R. Buckminster Fuller and his science of "Synergetics" had the answer in his "octet truss" (which he called an "isotrophic vector matrix", and which had been independently discovered about 50 years earlier by Alexander Graham Bell). You remember Fuller, right? The fellow who invented the geodesic dome?

Each of the struts composing the octet truss are the same length. Geometrically it is an array of tetrahedrons and octahedrons (in terms of Dungeons and Dragons polyhedral dice it uses d4's and d8's).

Sometimes instead of an octet truss designers will opt for a weaker but easier to construct space frame. The truss of the International Space Station apparently falls into this category.


The artifact was the shell of a solid fuel rocket motor. Part of the Mariner XX, from the lettering.

The Mariner XX, the ancient Pluto fly-by. Ages ago the ancient empty shell must have drifted back toward the distant sun, drifted into the thin Trojan-point dust and coasted to a stop. The hull was pitted with dust holes and was still rotating with the stabilizing impulse imparted three generations back.

As a collector's item the thing was nearly beyond price. Brennan took phototapes of it in situ before he moved in to attach himself to the flat nose and used his jet backpac to stop the rotation. He strapped it to the fusion tube of his ship, below the lifesystem cabin. The gyros could compensate for the imbalance.

In another sense the bulk presented a problem.

He stood next to it on the slender metal shell of the fusion tube. The antique motor was half as big as his mining singleship, but very light, little more than a metal skin for its original shaped-core charge. If Brennan had found pitchblende the singleship would have been hung with cargo nets under the fuel ring, carrying its own weight in radioactive ore. He would have returned to the Belt at half a gee. But with the Mariner relic as his cargo he could accelerate at the one gee which was standard for empty singleships.

There are few big cargo ships in the Belt. Most miners prefer to haul their own ore. The ships that haul large cargoes from asteroid to asteroid are not large; rather, they are furnished with a great many attachments. The crew string their payload out on spars and rigging, in nets or on lightweight grids. They spray foam plastic to protect fragile items. spread reflective foil underneath to ward off hot backlighting from the drive flame, and take off on low power.

The Blue Ox was a special case. She carried fluids and fine dusts; refined quicksilver and mined water, grain, seeds, impure tin scooped molten from lakes on dayside Mercury, mixed and dangerous chemicals from Jupiter's atmosphere. Such loads were not always available for hauling. So the Ox was a huge tank with a small threeman lifesystem and a fusion tube running through her long axis; but, since her tank must sometimes become a cargo hold for bulky objects, it had been designed with mooring gear and a big lid.

Nilsson's own small, ancient mining ship had become the Ox's lifeboat. The slender length of its fusion tube, flared at the end, stretched almost the length of the hold. There was an Adzhubei 4-4 computer, almost new; there were machines intended to serve as the computer's senses and speakers, radar and radio and sonics and monochromatic lights and hi-fi equipment. Each item was tethered separately, half a dozen ways, to hooks on the inner wall.

Nilsson nodded, satisfied, his graying blond Belter crest brushing the crown of his helmet. "Go ahead, Nate."

Nathan La Pan began spraying fluid into the tank. In thirty seconds the tank was filled with foam which was already hardening.

"Close 'er up."

Perhaps the foam crunched as the great lid swung down. The sound did not carry. Patroclus Port was in vacuum, open beneath the black sky.

The captive ship was small. Phssthpok found little more than a cramped life support system, a long drive tube, a ring-shaped liquid hydrogen tank with a cooling motor. The toroidal fuel tank was detachable, with room for several more along the slender length of the drive tube. Around the rim of the cylindrical life support system were attachments for cargo, booms and folded fine-mesh nets and retractable hooks.

He did find inspection panels in the drive tube. Within an hour he could have built his own crystal-zinc fusion tube, had he the materials. He was impressed. The natives might be more intelligent than he had guessed, or luckier. He moved up to the lifesystem and through the oval door.

The cabin included an acceleration couch, banks of controls surrounding it in a horseshoe, a space behind the couch big enough to move around in, an automatic kitchen that was part of the horseshoe, and attachments to mechanical senses of types frequently used in Pak warfare. But this was no warship. The natives' senses must be less acute than Pak senses. Behind the cabin were machinery and tanks of fluid, which Phssthpok examined with great interest.

One thing he understood immediately.

He was being very careful with the instrument panel. He didn't want to wreck anything before he found out how to pull astronomical data from the ship's computer. When he opened the solar storm warning to ascertain its purpose, he found it surprisingly small. Curious, he investigated further. The thing was made with magnetic monopoles.

From PROTECTOR by Larry Niven (1973)

A bit more simplistic is a simple stack of octahedrons (Dungeons and Dragons d8 polyhedral dice). This was used for the spine of the Valley Forge from the movie Silent Running (1972), later reused as the agro ship from original Battlestar Galactica.

Saddle Truss

Spacecraft spines are generally down the center of the spacecraft following the ship's thrust axis (the line the engine's thrust is applied along, usually from the center of the engine's exhaust through the ship's center of gravity).

This can be a pain to spacecraft designers if they have anything that needs to be jettisoned. Such items will have to be in pairs on opposite sides of the spine, and jettisoned in pairs as well. Otherwise the spacecraft's center of gravity will shift off the thrust axis, and the next time the engines are fired up it's pinwheel time.

In a NASA study TM-1998-208834-REV1 they invent a clever way to avoid this: the Saddle Truss.

The truss is a hollow framework cylinder with a big enough diameter to accommodate standard propellant tanks, consumables storage pods, and auxiliary spacecraft. One side of the cylinder frame is missing. The thrust axis is cocked a fraction of a degree off-center to allow for the uneven mass distribution of the framework.

The point is that tanks and other jettison-able items no longer have to be in pairs if you use a saddle truss. When it is empty you just kick it out through the missing side of the saddle truss. No muss, no fuss, and no having to have double the amount of propellant plumbing and related items.

Examples of the saddle truss can be found in the Bimodal NTR and the SNRE Spacecraft.

Waterskiing Spacecraft

This is a quite radical method to drastically reduce the structural mass of a spacecraft, allowing a handsome increase in valuable payload mass. It also dramatically increase the separation between a dangerously radioactive propulsion system and the crew, allowing a drastic decrease in the radiation shadow shield mass. This allows yet more handsome increases in valuable payload mass. As the cherry on top of the cake, it allows using the tumbling pigeon method of spin gravity without the direction of gravity inverting.

Please note this has never actually been used in a serious nuclear spacecraft design due to its unorthodox nature.

And warships with such a design would have their manoeuvring critically handicapped (or it's "crack-the-whip" time and the cable breaks).

The concept comes from the observation that for a given amount of structural strength, a compression member (such as a girder) generally has a higher mass that a corresponding tension member (such as a cable). And we know that every gram counts.

Charles Pellegrino and Dr. Jim Powell put it this way: current spacecraft designs using compression members are guilty of "putting the cart before the horse". At the bottom is the engines, on top of that is the thrust frame, and on top of that is rest of the spacecraft held together with girders (compression members) like a skyscraper. But what if you put the engine at the top and have it drag the rest of the spacecraft on a long cable (tension member). You'll instantly cut the structural mass by an order of magnitude or more!

And if the engines are radioactive, remember that crew radiation exposure can be cut by time, shielding, or distance. The advantage of distance is it takes far less mass than a shield composed of lead or something else massive. The break-even point is where the mass of the boom or cable is equal to the mass of the shadow shield. But the mass of a shadow shield is equal to the mass of a incredibly long cable. The HELIOS cable was about 300 to 1000 meters, the Valkyrie was ten kilometers.

But keep in mind that this design has no maneuverability at all. Agile it ain't. If you turn the ship too fast it will try to "crack the whip" and probably snap the cable. This probably makes the design unsuitable for warships, who have to jink a lot or be hit by enemy weapons fire.

Examples include HELIOS, the Valkyrie Antimatter Starship, the ramrobots from Larry Niven's A Gift From Earth and the ISV Venture Star from the movie Avatar.

Certain propulsion systems incorporate the waterskiing concept in spacecraft that use the propulsion. The main one is the Medusa, which sets off nuclear explosions inside a huge parachute-shaped sail. The sail accelerates, and drags along the payload on a long cable. Long because the payload does not want to be any closer to a series of nuclear explosions than it has to be.

The various types of sail propulsion drag the payload with a long cable as well. But for them, the long cable is not because the sail is radioactive, just that it is typically several kilometers in radius.

Cosine Thrust Loss

If the exhaust is radioactive or otherwise dangerous to hose the rest of the spacecraft with, you can have two or more engines angled so the plumes miss the ship.

Angled engines do reduce the effective thrust by an amount proportional to the cosine of the angle but for small angles it is acceptable. The delta V of the spacecraft is also reduced by the same proportion.

Note in the HELIOS design Krafft Ehricke figured that the 300 meter separation was enough to render the exhaust harmless so it does not angle the engine at all. Krafft has a single engine blasting straight at the habitat module. The only concession to the exhaust is mounting the cables on outriggers, so the cables do not pass through the zillion degree nuclear fireball exhaust plume. It would be most embarassing if the cables melted.


The HELIOS has a thrust of 981,000 newtons. Say that Dr. Ehricke figured the exhaust would be dangerous to the habitat module, so the single engine would have to be replaced by two engines with 490,500 newtons each angled off-center by 5°. What would that do to the thrust?

CosineFactor = cos(OffAngle)
CosineFactor = cos(5°)
CosineFactor = 0.9962…

So both the thrust and delta V would be at 99.62%

Each engine has a thrust of 490,500 newtons, this would be reduced by the cosine thrust loss to:

EffectiveThrust = ActualThrust * CosineFactor
EffectiveThrust = 490,500 * 0.9962
EffectiveThrust = 488,636 newtons

488,636 newtons * 2 engines = 977,272 newtons. This means that angling the engines lowers the total thrust by 3,728 newtons, which is an ouch but not a show stopper. If the thrust absolutely has to be 981,000 newtons total, each engine would have to have its thrust increased from 490,500 to 492,371 newtons in order to compensate for the cosine loss.

If the HELIOS had a delta-V of 21,000 m/s, the cosine loss would reduce it to 21,000 * 0.9962 = 20,920 m/s. This is a loss of 80 m/s which is not negligible but not a show-stopper either.

Here's how we can shave off many tons of shielding.

Put the engine up front and carry the crew compartment ten kilometers behind the engine, on the end of a tether. Let the engine pull the ship along, much like a motorboat pulling a water skier, and let the distance between the gamma ray source and the crew compartment, as the rays stream out in every direction, provide part of the gamma ray protection - with almost no weight penalty at all. (ed. note: this should remind you of "Helios") We can easily direct the pion/muon thrust around the tether and its supporting structures, and we can strap a tiny block of (let's say) tungsten to the tether, about one hundred meters behind the engine. Gamma rays are attenuated by a factor of ten for every two centimeters of tungsten they pass through. Therefore, a block of tungsten twenty centimeters deep will reduce the gamma dose to anything behind it by a factor of ten to the tenth power (1010). An important shielding advantage provided by a ten-kilometer-long tether is that, by locating the tungsten shield one hundred times closer to the engine than the crew, the diameter of the shield need be only one-hundredth the diameter of the gamma ray shadow you want to cast over and around the crew compartment. The weight of the shielding system then becomes trivial.

The tether system requires that the elements of the ship must be designed to climb "up" and "down" the lines, somewhat like elevators on tracks.

We can even locate the hydrogen between the tungsten shadow shield and the antihydrogen, to provide even more shielding for both the crew and the antihydrogen.

There is an irony involved in this configuration. Our "inside-out" rocket, the most highly evolved rocket yet conceived, is nothing new. We have simply come full circle and rediscovered Robert Goddard's original rocket configuration: with engines ahead of the fuel tanks and the fuel tanks ahead of the payload.

From FLYING TO VALHALLA by Charles Pellegrino (1993)

Interstellar Ramscoop Robot #143 left Juno at the end of a linear accelerator. Coasting toward interstellar space, she looked like a huge metal insect, makeshift and hastily built. Yet, except for the contents of her cargo pod, she was identical to the last forty of her predecessors. Her nose was the ramscoop generator, a massive, heavily armored cylinder with a large orifice in the center. Along the sides were two big fusion motors, aimed ten degrees outward, mounted on oddly jointed metal structures like the folded legs of a praying mantis. The hull was small, containing only a computer and an insystem fuel tank.

(ed note: cosine of 10° is about 0.9848, so thrust and delta V would be reduced to 98.5%.)

Juno was invisible behind her when the fusion motors fired. Immediately the cable at her tail began to unroll. The cable was thirty miles long and was made of braided Sinclair molecule chain. Trailing at the end was a lead capsule as heavy as the ramrobot itself.

(ed note: "Sinclair molecule chain" is an unobtainium wire that is only one molecule thick and absurdly strong. The theoretical ultimate of low mass cable.)

On twin spears of actinic light the ramrobot approached Pluto's orbit. Pluto and Neptune were both on the far side of the sun, and there were no ships nearby to be harmed by magnetic effects.

The ramscoop generator came on.

The conical field formed rather slowly, but when it had stopped oscillating, it was two hundred miles across. The ship began to drag a little, a very little, as the cone scooped up interstellar dust and hydrogen. She was still accelerating. Her insystem tank was idle now, and would be for the next twelve years. Her food would be the thin stuff she scooped out between the stars.

From A GIFT FROM EARTH by Larry Niven (1968)

Space Trains and Truckers

As attractive as is the admirable reduction in radiation shield mass offered by the waterskiiing spacecraft concept, there are practical problems in being towed on the end of a kilometer-long cable.

But the bit about using tension instead of compression members is still a worthwhile idea. Take some species of space tug, mount the propusion system on outriggers so the exhaust does not fire directly back along the ship's spine, and attach cargo modules to massive couplings on the bottom of the thrust frame.

If the space tug is hauling one or two cargo modules, this resembles an 18-wheeler hauling a couple of semi-trailers. A "space-trucker" so to speak.

However, if the tug is hauling a long chain of cargo modules, this is more like a freight train with a locomotive at the front and a string of freight cars in tow. A "space train" as it were.

As with all cargo spacecraft, delicate cargo will be housed in pressurized temperature-controlled cannisters but bulk ores and other insensitive cargo will just be dragged along in nets. And much like railroad locomotives, this will be less like loading crates into a seagoing container ship and more like latching cargo cans into long strings terminating at the rocket engine at the top.




“There it is,” Shadow Jack said, with almost a sigh. “Mecca rock.”

Betha watched it come into view at the port: a fifty-kilometer potato-shaped lump of stone, scarred by nature’s hand and man’s. Mecca’s long axis pointed to the sun; the side nearest them lay in darkness, haloed by an eternal corona of sunglare. As they closed she began to see landing lights; and, between them, immense shining protrusions lit from below, throwing their shadows out to be lost in the shadow of the void. She identified them finally as storage tanks—enormous balloons of precious gases. At last… She stirred in the narrow, dimly lit space before the instruments, felt her numbed emotions stir and come alive.

“Out there, Shadow Jack.” She leaned closer to the port, rubbed the fog of moisture from the glass. “A tanker coming in.”

He peered past her. They saw the ship, still lit by the sun: a ponderous metallic tick, its plastic belly bloated with precious gases and clutched inside three legs of steel, booms for the ship’s nuclear-electric rockets. “Look at the size of that! It must be comin’ in from the Rings. They wouldn’t use that on local hauls." He raised his head, following its downward arc. “Down there, that must be the docking field.”

She could see the field clearly now, an unnatural gleaming smoothness in the artificial/light, cluttered with cranes and ringed by more mechanical parasites, gorged and empty. Smaller craft moved above them, fireflies showing red: sluggish tows in a profusion of makeshift incongruity.

From THE OUTCASTS OF HEAVEN BELT by Joan Vinge (1978)

He glanced up, at the purity of blackness unmarred by atmosphere, at the stars. Somewhere below his feet, through kilometers of nearly solid rock, was the tiny, pale spinel of the sun Heaven. He would be seeing it again, soon enough—he focused on the looming grotesqueness tethered at the end of the mooring cable, bifurcated by the abrupt edge of the asteroid’s horizon: the converted volatile freighter that would take them across the Main Belt and on in to Heaven’s second planet, to pick up one man…and a treasure. The three jutting booms that kept its nuclear electric rockets suspended away from the living quarters clutched rigid cylinders instead of the usual flimsy volatiles sack; it carried a liquid fuel rocket for their descent to the planet’s surface.

From LEGACY by Joan Vinge (1980)

(ed note: this is a laser thermal rocket energized by powerful remote lasers on an L5 colony)

     The next item on the agenda was the laser-powered high-acceleration tug, otherwise referred to as the ultra-fast optical system, UFOS having more dash and elan than LPHAT. Corporate Susan made the presentation it had worked up with Skaskash and Lady Dark.

     "The basic idea isn't bad,” said Cantrell. “How would you keep the lens oriented normal to the laser when you start to move the engine to a different orientation?"
     "We have a pair of pipes at the equator of the sphere, pumping water in opposite directions,” said Corporate Susan. “Also, inside the sphere, under the photovoltaic surface, are two pairs of circular loops, set flush with the surface and at right angles to each other. Each pair pumps water in a counterrotary direction. The pumps are all controlled, so the UFOS is gyroscopically stabilized in three planes."

     "I see,” Dornbrock said. “How do you move the engine around on the surface of the geodesic sphere?"
     "The sphere rests on this little egg cup here,” said Corporate Susan. “The egg cup is a plastic perforated surface. When we want to move, we pressurize the surface, and the geodesic sphere floats on an air cushion. Then the mechanical hands around the perimeter of the egg cup orient the engine while the sphere stays put, or the engine stays put and the hands reorient the sphere, depending on how you work the gyroscopic pumps."
     "Wouldn't you lose a lot of air pressurizing the perforated surface?” Corporate Forziati asked.
     "No, actually,” Corporate Susan replied. “We have built a little valve into each perforation which only operates when the surface is depressed by the weight of the element of the sphere in contact with it.” A diagram flashed on her telecon screen for a moment.
     "Thank you,” Forziati said. “And when you are not under thrust, weight is no problem and you don't pressurize. Very good."

     "On the other end of the egg cup,” Bogdanovitch said, “where you have the engines and the tanks for the reaction mass, you have a long cable supporting the warship. Couldn't you have the ship on an egg cup, too?"
     "No,” Corporate Susan replied. “The engines are thrusting against the geodesic sphere, which rests on top of the egg cup. The warship must keep its center of mass in line with the axis of thrust. Put it on the sphere with its own egg cup, and it would have to stay lined up with the engines—on the other side—which means the sphere would have to be built stronger, and heavier."
     "And it would get in the way of the big laser beam,” Cantrell added.

     "Then how does the warship stay out of the jet of ions?” asked Bogdanovitch.
     "It rotates at the end of its cable,” said Corporate Susan, “and makes a little circle around the jet of uranium ions which provide the main thrust. The jet of boron and hydrogen is flared off, simply to provide electrical neutrality, but it also provides a tiny bit of thrust, which can be used to offset the wobble the ship would otherwise cause by swinging around the main jet."

     "I don't understand,” Marian said.
     Corporate Susan dissolved into a diagram. “Consider the vector diagram of the force exerted by the cable supporting the ship,” said the computer. “Most of it runs through the axis of thrust, but there is a small component going at right angles to that thrust. The boron and hydrogen, flared off with the excess electrons from the decaply ionized uranium, can be adjusted to exactly balance that small component. The flare—a very soft jet—would be in the same plane as the ship, and pointing in the same direction, to push where the ship is pulling."
     "The jet—the flare, I mean, turns with the cable?” asked Marian.
     "Of course,” said Corporate Susan.

     "Orange and green,” said Marian. “Very pretty. What color is the uranium jet?"
     "Hard X-ray,” Skaskash said. “It would probably be dangerous for two or three hundred kilometers."
     "That might be an idea whose time has come,” Cantrell said at last. “Any more questions? No? Shall we build it? ... It seems to be unanimous."

     "I have a model at the shop you can use. I'll have the changes you wanted put on, and you can use that, if you want."
     "How big is it?"
     "Not big—” Ilgen stretched his arms. “Maybe a meter and a half. Did you decide what ship you wanted with it?"
     "The Alamo. We need to impress people, and the Alamo is the biggest thing we've got."
     "Right, Charlie. I'll throw in a model of the Alamo to the same scale. The UFOS plus the Alamo figures to go seven to eight times as fast as any cruiser."
     Cantrell whistled softly.
     "I'll tell the Navy,” he said. “That ought to make them very happy."

From THE PIRATES OF ROSINANTE by Alexis Gilliland (1982)

When space westerns decide to get literal with the genre name, these guys tend to show up.

They are usually depicted as, well, southern-fried semi truckers that happen to fly cargo spaceships instead. Usually easygoing, unflappable, and have a backwoods wisdom developed during the time spent alone on long hauls between solar systems. Occasionally they're smugglers, but usually aren't drug mules. From time to time, between hauls they'll stop at a local tavern and challenge various folks to a little arm-wrestling, drinking contests and then drunken brawls. Tend to talk in a Southern or slight Texan drawl.

Overlaps occasionally with the Boisterous Bruiser, Gentle Giant, Warrior Poet or other ersatz cowboy-style archetypes.

Their ships are most likely Used Future pieces of junk, just barely held together — bonus points if they're blocky, and bear a surprising resemblance to modern Mack trucks or other 18-wheelers.

Often a subtrope of Intrepid Merchant.

(ed note: see TV Trope page for list of examples)

We had a lot of luck on Venus
We always had a ball on Mars
Meeting all the groovey people
We've rocked the Milky Way so far
We danced around with Borealis
We're space truckin' round the the stars
Come on let's go Space Truckin'

Remember when we did the moonshot
And Pony Trekker led the way
We'd move to the Canaveral moonstop
And everynaut would dance and sway
We got music in our solar system
We're space truckin' round the stars
Come on let's go Space Truckin'

The fireball that we rode was moving
But now we've got a new machine
Yeah Yeah Yeah Yeah the freaks said
Man those cats can really swing
They got music in their solar system
They've rocked around the Milky Way
They dance around the Borealis
They're Space Truckin' everyday
Come on
From SPACE TRUCKIN' by Deep Purple (1972)




This was the lead ship for the warp superconvoys, the 100-kilometer cargo carriers that revolutionized interstellar industrial transport. Configured in 8-ship linked octogons at the head of the convoy, with 4-ship squares of booster tugs after each 10-container segment, and all controls subspace-radio synchronized, these superconvoys transported billions of kilograms per superconvoy.

Length225 m
Beam220 m
Draught45.6 m
Mass72.5 million kg
Ship's Complement
Std Ship's Complement66
Range2000 light-years
Cruising SpeedEmpty - Warp 3.5
Loaded - Warp 2
EnginesAdv 3rd Gen Warp Drive
Fuel1:1 matter/antimatter
NotesTractor beam coupling for cargo containers
Most powerful thrust in starship history

SUPERCONVOYS OPEN NEW ERA OF TRADE: Billion ton ships are boon to industry (2161)

Engineering Log:

It's working! Warp effect is being engaged, and this superconvoy is rolling! Next stop—Centauri Spaceworks.

I'll let the boys upstairs take the glory, but the truth is old greasemonkey Sabella down here in the Engine Room is who straightened this whole mess out, I must admit in all humility.

Problem: how to transport raw materials from whistle-stop asteroid belts in the boondock sectors to the space factories of the UFP? And I'm not talking a freighter or two here. I'm talking about the billions and billions of metric tonnage needed each stardate to feed the Federation's ravening industrial plant. Star Fleet is stymied.

Solution: Sabella to the rescue. lt's elementary, I patiently explain, what we need are warp convoys of a hundred freighters and more. It can't be done, sneer the design baboons. No ship could produce a warp effect that great. Who said anything about one ship? I reply.

So I spelled it out for them. At the head of the convoy, assemble a configuration of heavy warp-drive tugs, say eight of them in an octagon. Lock their controls together and use the whole pile as the inertial driver. Next, string out a dozen or so of those new kilometer-long cargo cans, coupling them with tractor beams. Then—and here my brilliance staggers even my own modest self—plug in another configuration of warp tugs, four should do it, and knit the rest of the convoy with the same pattern, synchronize it all with subspace radio so that all warp engines engage simultaneously, and ride into the sky!

It can't be done, cry the designers.

An interesting conjecture, muses star Fleet.

It can be done, admit the designers in wonderment. Maybe only warp 2 or so, but this would save years of travel time, and trillions of credits.

Trillions? asks Star Fleet. Maybe we'll consider it. They considered it, they did it, it's done.

Merely brilliant, Sabella, I say. Merely brilliant.


Admiral's Log:

Our flagship Endurance has run a final sensor scan of Bayard’s Planet, and I can report with certainty that not a single inhabitant has been missed by this, the largest spacelift in history. None too soon either. The shock wave from the supernova explosion that destroyed the Kepler will reach this systems sun in less than three star-dates, and it is 100 percent probable that it will initiate a chain reaction nova.

And the nova will destroy all life in the star system. All oceans will boil away; the gamma ray bombardment will scorch the surface; the planet itself may undergo an internal disruption and blow apart. it was imperative that Bayard’s Planet be evacuated. But how to evacuate ten million inhabitants? With the biggest spacelift ever.

Every starship within a fifty light-year radius was pressed into the effort by special request of the UFP. And I am just amazed by the response. Cruisers, destroyers, scout-ships, corvettes, all rendezvoused. Also trading clippers, low-warp tugs, even pleasure crafts of every type all rallied to the cause.

But all these valiant volunteers twice over could not have removed ten million people in time. Then an engineer in Star Fleet Merchant Marine named Sabella came up with the novel idea of refitting those new super-convoys with life support and jury-rigged decks. And that turned the corner for us. Starbase 5 renovated two superconvoys to accommodate passengers in an amazingly short time, and after weeks of impossible logistics and unmeetable timetables, we have succeeded in evacuating the entire planet. The galaxy has never seen an operation of this scope before. And I for one fervently hope it never has to again.

From STAR TREK SPACEFLIGHT CHRONOLOGY by Stanley and Fred Goldstein (1980)

Thermal Protection


Multi-layer insulation, or MLI, is thermal insulation composed of multiple layers of thin sheets and is often used on spacecraft. It is one of the main items of the spacecraft thermal design, primarily intended to reduce heat loss by thermal radiation. In its basic form, it does not appreciably insulate against other thermal losses such as heat conduction or convection. It is therefore commonly used on satellites and other applications in vacuum where conduction and convection are much less significant and radiation dominates. MLI gives many satellites and other space probes the appearance of being covered with gold foil.

Function and design

The principle behind MLI is radiation balance. To see why it works, start with a concrete example - imagine a square meter of a surface in outer space, held at a fixed temperature of 300 K, with an emissivity of 1, facing away from the sun or other heat sources. From the Stefan–Boltzmann law, this surface will radiate 460 W. Now imagine placing a thin (but opaque) layer 1 cm away from the plate, thermally insulated from it, and also with an emissivity of 1. This new layer will cool until it is radiating 230 W from each side, at which point everything is in balance. The new layer receives 460 W from the original plate. 230 W is radiated back to the original plate, and 230 W to space. The original surface still radiates 460 W, but gets 230 W back from the new layers, for a net loss of 230 W. So overall, the radiation losses from the surface have been reduced by half by adding the additional layer.

More layers can be added to reduce the loss further. The blanket can be further improved by making the outside surfaces highly reflective to thermal radiation, which reduces both absorption and emission.

The layers of MLI can be arbitrarily close to each other, as long as they are not in thermal contact. The separation space only needs to be minute, which is the function of the extremely thin scrim or polyester 'bridal veil' as shown in the photo. To reduce weight and blanket thickness, the internal layers are made very thin, but they must be opaque to thermal radiation. Since they don't need much structural strength, these internal layers are usually made of very thin plastic, about 6 μm (1/4 mil) thick, such as Mylar or Kapton, coated on one side with a thin layer of metal on both sides, typically silver or aluminium. For compactness, the layers are spaced as close to each other as possible, though without touching, since there should be little or no thermal conduction between the layers. A typical insulation blanket has 40 or more layers. The layers may be embossed or crinkled, so they only touch at a few points, or held apart by a thin cloth mesh, or scrim, which can be seen in the picture above. The outer layers must be stronger, and are often thicker and stronger plastic, reinforced with a stronger scrim material such as fiberglass.

In satellite applications, the MLI will be full of air at launch time. As the rocket ascends, this air must be able to escape without damaging the blanket. This may require holes or perforations in the layers,[2] even though this reduces their effectiveness.

MLI blankets are constructed with sewing technology. The layers are cut, stacked on top of each other, and sewn together at the edges. Seams and gaps in the insulation are responsible for most of the heat leakage through MLI blankets. A new method is being developed to use polyetheretherketone (PEEK) tag pins (similar to plastic hooks used to attach price tags to garments) to fix the film layers in place instead of sewing to improve the thermal performance.

Additional properties

Spacecraft also may use MLI as a first line of defense against dust impacts. This normally means spacing it a cm or so away from the surface it is insulating. Also, one or more of the layers may be replaced by a mechanically strong material, such as beta cloth.

In some applications the insulating layers must be grounded, so they cannot build up a charge and arc, causing radio interference. Since the normal construction results in electrical as well as thermal insulation, these applications may include aluminum spacers as opposed to cloth scrim at the points where the blankets are sewn together.

From the Wikipedia entry for MULTI-LAYER INSULATION


There are some hazards to worry about with these space-age materials. Titanium and magnesium are extremely flammable (in an atmosphere containing oxygen). And when I say "extremely" I am not kidding.

Do not try to put out a magnesium fire by throwing water on it. Blasted burning magnesium will suck the oxygen atoms right out of the water molecules, leaving hydrogen gas (aka what the Hindenburg was full of). A carbon-dioxide fire extinguisher won't work either, same result as water except you get a cloud of carbon instead of hydrogen. Instead use a Class D dry chemical fire extinguisher or a lot of sand to cut off the oxygen supply. Oh, did I mention that burning magnesium emits enough ultraviolet light to permanently damage the retinas of the eyes?

The same goes for burning titanium. Except there is no ultraviolet light, but there is a chance of ignition if titanium is in contact with liquid oxygen and the titanium is struck by a hard object. It seems that the strike might create a fresh non-oxidized stretch of titanium surface, which ignites the fire even though the liquid oxygen is at something like minus 200° centigrade. This may mean that using titanium tanks for your rocket's liquid oxygen storage is a very bad idea.

An emergency crew at a spaceport, who has to deal with a crashed rocket, will need the equipment to deal with this.

And if the titanium, magnesium, or aluminum becomes powdered, you have to stop talking in terms of "fire" and start talking in terms of "explosion."


As an interesting side note, rockets constructed of aluminum are extremely vulnerable to splashes of metallic mercury or dustings of mercury salts. On aluminum, mercury is an "oxidizing catalyst", which means the blasted stuff can corrode through an aluminum beam in a matter of hours (in an atmosphere containing oxygen, of course). This is why mercury thermometers are forbidden on commercial aircraft.

Why? Ordinarily aluminum would corrode much faster than iron. However, iron oxide, i.e., "rust", flakes off, exposing more iron to be attacked. But aluminum oxide, i.e., "sapphire", sticks tight, protecting the remaining aluminum with a gem-hard barrier. Except mercury washes the protective layer away, allowing the aluminum to be consumed by galloping rust.

Alkalis will have a similar effect on aluminum, and acids have a similar effect on magnesium (you can dissolve magnesium with vinegar). As far as I know nothing really touches titanium, its corrosion-resistance is second only to platinum.

Protective Paint

If you want a World War II flavor for your rocket, any interior spaces that are exposed to rain and other corrosive planetary weather should be painted with a zinc chromate primer. Depending on what is mixed into the paint, this will give a paint color ranging from yellowish-green to greenish-yellow. In WWII aircraft it is found in wheel-wells and the interior of bomb bays. In your rocket it might be found on landing jacks and inside airlock doors.

Naturally this does not apply to strict orbit-to-orbit rockets, or rockets that only land on airless moons and planets. Well, now that I think about it, some of the lunar dust is like clouds of microscopic razor blades so they are dangerously abrasive.


Zinc chromate, ZnCrO4, is a chemical compound containing the chromate anion, appearing as odorless yellow powder or yellow-green crystals, but, when used for coatings, pigments are often added. It is used industrially in chromate conversion coatings, having been developed by the Ford Motor Company in the 1920s.


Zinc chromate’s main use is in industrial painting as a coating over iron or aluminum materials. It was used extensively on aircraft by the U.S. military, especially during the 1930s and 1940s, but is also used in a variety of paint coatings for the aerospace and automotive industries. Its use as a corrosion-resistant agent was applied to aluminium alloy parts first in commercial aircraft, and then in military ones. During the 1940 and 1950s it was typically found as the "paint" in the wheel wells of retractable landing gear on U.S. military aircraft to protect the aluminium from corrosion. This compound was a useful coating because it is an anti-corrosive and anti-rust primer. Since it is highly toxic it also destroys any organic growth on the surface. Zinc chromate is also used in spray paints, artists’ paints, pigments in varnishes, and in making linoleum.


Recent studies have shown that not only is zinc chromate highly toxic, it is also a carcinogen. Exposure to zinc chromate can cause tissue ulceration and cancer. A study published in the British Journal of Industrial Medicine showed a significant correlation between the use of zinc chromate and lead chromate in factories and the number of cases in lung cancer experienced by the workers. Because of its toxicity the use of zinc chromate has greatly diminished in recent years.

From the Wikipedia entry for ZINC CHROMATE

Rocket Tumble

The basic idea is that the Axis of Thrust from the engines had better pass through the the spacecraft's center of gravity (CG) or everybody is going to die. In addition, if the spacecraft is currently passing through a planet's atmosphere the axis of thrust had better be parallel to the aerodynamic axis or the same thing will happen.

Specifically, "everybody is going to die" means the spacecraft is going to loop-the-loop or tumble like a cheap Fourth-of-July skyrocket (Heinlein calls this a rocket "falling off its tail"). If this happens during lift-off the ship will auger into the ground like a nuclear-powered Dinosaur-Killer asteroid and make a titanic crater. If it happens in deep space, the rocket will spin like a pinwheel firework spraying atomic flame everywhere. This will waste precious propellant, give the spacecraft a random vector, and severely injure the crew with unexpected spin gravity. If they are lucky the crew's broken bones will heal about the same time that they run out of oxygen.

The axis of thrust is a line starting at the center of the exhaust nozzle's throat, and traveling in the exact opposite direction of the hot propellant. It is the direction that the thrust is pushing the rocket. As long as the axis of thrust passes through the CG, the spacecraft will be accelerated in that direction. If the axis of thrust is not passing through the CG, the spacecraft will start to spin around the CG. When done on purpose this is called a yaw or pitch maneuver. When this is done by accident, it is called OMG WE'RE ALL GOING TO DIE!

Some engines can be gimbaled, rotating their axis of thrust off-center by a few degrees. This is intended for yaw and pitch, but it can be used in emergencies to cope with accidental changes in the center of gravity (e.g., the cargo shifts).

When laying out the floor plan, you want the spacecraft to balance. This boils down to ensuring that the ship's center of gravity is on the central axis, which generally is the same as the axis of thrust. There are exceptions. The Grumman Space Tug has its center of gravity shift wildly when it jettisons a drop tank. To compensate, the engine can gimbal by a whopping ±20°.

Balancing also means that each deck should be "radially symmetric". That's a fancy way of saying that if you have something massive in the north-west corner of "D" deck, you'd better have something equally massive in the south-east corner. Otherwise the center of gravity won't be centered.

This is another reason to strap down the crew during a burn. Walking around could upset the ship's balance, resulting in the dreaded rocket tumble. This will be more of a problem with tiny ships than with huge cruisers, of course. The same goes for the cargo. The load-master better be blasted sure all the tons of cargo are nailed down so they don't shift. And be sure the cargo is evenly balanced around the ship's axis to keep the center of gravity in the center.

Small ships might have "trim tanks", small tanks into which water can be pumped in order to adjust the balance. The ship will also have heavy gyroscopes that will help prevent the ship from falling off its tail, but there is a limit to how much imbalance that they can compensate for.

Propellent Tankage

A cursory look at the rocket's mass ratio will reveal that most of the rocket's mass is going to be propellant tanks.

For anything but a torchship, the spacecraft's mass ratio is going to be greater than 2 (i.e., 50% or more of the total mass is going to be propellant). Presumably the propellant is inside a propellant tank (unless you are pulling a Martian Way gag and freezing the fuel into a solid block). Remember, RockCat said all rockets are giant propellant tanks with an engine on the bottom and the pilot's chair at the top.

If you have huge structure budget, you have a classic looking rocket-style rocket with propellant tanks inside. If you have a medium structure budget, you have a spine with propellant tanks attached. If you have a small structure budget, you'll have an isogrid propellant tank for a spine, with the rest of the rocket parts attached.

And if you are stuck with a microscopic structure budget, you'll have a foil-thin propellant tank stiffened by the pressure of the propellant, with the rest of the rocket parts attached. But the latter tends to collapse when the propellant is expended and the pressure is gone. This was used in the old 1957 Convair Atlas rocket, but not so much nowadays. You cannot really reuse them.


It had all seemed perfectly logical back on Mars, but that was Mars. He had worked it out carefully in his mind in perfectly reasonable steps. He could still remember exactly how it went. It didn't take a ton of water to move a ton of ship. It was not mass equals mass, but mass times velocity equals mass times velocity. It didn't matter, in other words, whether you shot out a ton of water at a mile a second or a hundred pounds of water at twenty miles a second. You got the same velocity out of the ship.

That meant the jet nozzles had to be made narrower and the steam hotter. But then drawbacks appeared. The narrower the nozzle, the more energy was lost in friction and turbulence. The hotter the steam, the more refractory the nozzle had to be and the shorter its life. The limit In that direction was quickly reached.

Then, since a given weight of water could move considerably more than its own weight under the narrow-nozzle conditions, it paid to be big. The bigger the water-storage space, the larger the size of the actual travel-head, even in proportion. So they started to make liners heavier and bigger. But then the larger the shell, the heavier the bracings, the more difficult the weldings, the more exacting the engineering requirements. At the moment, the limit in that direction had been reached also.

And then he had put his finger on what had seemed to him to be the basic flaw—the original unswervable conception that the fuel had to be placed inside the ship; the metal had to be built to encircle a million tons of water.

Why? Water did not have to be water. It could be ice, and ice could be shaped. Holes could be melted into it. Travel-heads and jets could be fitted into it. Cables could hold travel-heads and jets stiffly together under the influence of magnetic field-force grips.

From THE MARTIAN WAY by Isaac Asimov (1952)

Our running example Polaris spacecraft has a gas core nuclear thermal rocket engine.

The fuel is uranium 235. It will probably be less than 1% of the total propellant load so we will focus on just the propellant for now.

Nuclear thermal rockets generally use hydrogen since you want propellant with the lowest molecular mass. Liquid hydrogen has a density of 0.07 grams per cubic centimeter.

The Polaris has 792.6 metric tons of hydrogen propellant. 792.6 tons of propellant = 792,600,000 grams / 0.07 = 11,323,000,000 cubic centimeters = 11,323 cubic meters . The volume of a sphere is 4/3πr3 so you can fit 11,323 cubic meters in a sphere about 14 meters in radius . Almost 92 feet in diameter, egad! It is a pity hydrogen isn't a bit denser.

If this offends your aesthetic sense, you'll have to go back and change a few parameters. Maybe a 2nd generation GC rocket, and a mission from Terra to Mars but not back. Maybe use methane instead of hydrogen. It only has an exhaust velocity of 6318 m/s instead of hydrogen's superior 8800 m/s, but it has a density of 0.42 g/cm3, which would only require a 1.7 meter radius tank. (Methane has a higher exhaust velocity than one would expect from its molecular weight, due to the fact that the GC engine is hot enough to turn methane into carbon and hydrogen. Note that in a NERVA style engine the reactor might become clogged with carbon deposits.)

Propellant Tank Mass

Robert Zubrin says that as a general rule, the mass of a fuel tank loaded with liquid hydrogen will be about 87% hydrogen and 13% tank. In other words, multiply the mass of the liquid hydrogen by 0.15 to get the mass of the empty tank (0.13 / 0.87 = 0.15).

So the Polaris' 792.6 tons of hydrogen will need a tank that masses 792.6 * 0.15 = 119 tons.

87% propellant and 13% tank is for a rocket designed to land on a planet or that is capable of high acceleration. An orbit-to-orbit rocket could get by with more hydrogen and less tank. This is because the tanks can be more flimsy since they will not have to endure the stress of landing (A landing-capable rocket that uses a propellant denser than hydrogen can also get away with a smaller tank percentage). Zubrin gives the following ballpark estimates of the tank percentage:

PropellentEngineTank %
ArgonIon rocket4
WaterNuclear salt water rocket4
HydrogenNTR / GCR10

But if you want to do this the hard way, you'd better warm up your slide rule.

The total tank volume (Vtot) of a tank is the sum of four components:

  1. Usable Propellant Volume (Vpu): the volume holding the propellant that can actually be used.
  2. Ullage Volume (Vull): the volume left unfilled to accomodate expansion of the propellant or contraction of the tank structure. Typically 1% to 3% of total tank volume.
  3. Boil-off Volume (Vbo): For cryogenic propellants only. The volume left unfilled to allow for the propellant that boils from liquid to gas due to external heat.
  4. Trapped Volume (Vtrap): the volume of unusable propellant left in all the feed lines, valves, and other components after the tank is drained. Typically the volume of the feed system.

Vtot = Vpu + Vull + Vbo + Vtrap

No, I do not know how to estimate the Boil-off Volume. A recent study estimated that in space cryogenic tanks suffered an absolutely unacceptable 0.1% boiloff/day, and suggested this had to be reduced by an order of magnitude or more. When the boil-off volume is full, a pressure relief valve lets the gaseous propellant vent into space, instead of exploding the tank.

Tanks come in two shapes: spherical and cylindrical. Spherical are better, they have the most volume for the least surface area, so are the lightest. But many spacecraft have a limit to their maximum diameter, especially launch vehicles. In this case cylindrical has a lower mass than a series of spherical tanks.

The internal pressure of the propellant has the greatest effect on the tank's structural requirements. Not as important but still significant are acceleration, vibration, and handling loads. Unfortunately I can only find equations for the effects of internal pressure. Acceleration means that tanks which are in high-acceleration spacecraft or in spacecraft that take-off and land from planets will have a higher mass than tanks for low-acceleration orbit-to-orbit ships. My source did say that figuring in acceleration, vibration, and handling would make the tank mass 2.0 to 2.5 times as large as what is calculated with the simplified equations below.

In the Space Shuttle external tank, the LOX tank was pressurized to 150,000 Pa and the LH2 tank was pressurized to 230,000 Pa.

The design burst pressure of a tank is:

Pb = fs * MEOP


Pb = design burst pressure (Pa)
fs = safety factor (typically 2.0)
MEOP = Maximum Expected Operating Pressure of the tank (Pa)

Tank Materials

Allowable Strength
Mass Factor
2219 - Aluminum2,8000.413
0.214 welded
4130 - Steel7,8300.86211.232,500
Graphite Fiber

Spherical Tanks

You have to make Vs so it is equal to Vtot, or at least equal to Vtot - Vtrap.

Vs = 4/3 * π * rs3

As = 4 * π * rs2

ts = (Pb * rs) / (2 * Ftu)

Ms = As * ts * ρ


rs = radius of sphere (m)
As = surface area of sphere (m2)
Vs = volume of sphere (m3)
ts = wall thickness of sphere (m)
Pb = design burst pressure (Pa)
Ftu = allowable material strength (Pa) from tank materials table
Ms = mass of spherical tank (kg)
ρ = density of tank structure material (kg/m3 from tank materials table

Cylindrical Tanks

Cylindrical tanks are cylinders where each end is capped with either hemispheres (where radius and height are equal) or hemiellipses (where radius and height are not equal). As it turns out cylindrical tanks with hemiellipses on the ends are always more massive than hemispherical cylindrical tanks. So we won't bother with the equations for hemielliptical tanks. In the real world rocket designers sometimes use hemielliptical tanks in order to reduce tank length.

What you do is calculate the mass of the cylindrical section of the tank Mc using the equations below. Then you calculate the mass of the two hemispherical endcaps (that is, the mass of a single sphere) Ms using the value of the cylindrical section's radius for the radius of the sphere in the spherical tank equations above. The mass of the cylindrical tank is Mc + Ms.

Vc = π * rc2 * lc

Ac = 2 * π * rc2 * lc

tc = (Pb * rc) / Ftu

Mc = Ac * tc * ρ


rc = radius of cylindrical section (m)
lc = length of cylindrical section (m)
Ac = surface area of cylindrical section (m2)
Vc = volume of cylindrical section (m3)
Pb = design burst pressure (Pa)
Ftu = allowable material strength (Pa) from tank materials table
ρ = density of tank structure material (kg/m3 from tank materials table
tc = wall thickness of cylindrical section (m)
Mc = mass of cylindrical tank section (kg)


When the rocket is sitting on the launch pad, the planet's gravity pulls the propellant down so that the pumps at the aft end of the tank can move it to the engine. When the rocket is under acceleration, the thrust pulls the propellant down to the pumps. Once the engines cut off and the rocket is in free fall, well, the remaining pooled at the bottom turns into zillions of blobs and starts floating everywhere. See the video:

This isn't a problem, up until the point where you want to start the engine up again. Trouble is, the propellant isn't at the aft pump, it is flying all over the place. What's worse, some of the liquid propellant might have turned into bubbles of gas, which could wreck the engine if they are sucked into the pump. Vapor lock in a rocket engine is an ugly thing.

In 1960 Soviet engineers invented the solution: Ullage Motors. These are tiny rocket engines that only have to accelerate the rocket by about 0.001g (0.01 m/s). That's enough to pull the propellant down to the pump, and to form a boundary between the liquid and gas portions. In some cases, the spacecraft's reaction control system (attitude jets) can operate as ullage motors.

In the Apollo service module, they use a "retention reservoir" instead of an ullage burn (but they have to burn anyway if the amount of fuel and oxidizer drops below 56.4%).

Liquid oxygen in the oxidizer storage tank flows into the oxidizer sump tank. During an engine burn, oxygen flows to the bottom of the sump tank, through an umbrella shaped screen, into the retention reservoir, then into a pipe at the bottom leading to the engine. The same system is used in the fuel tanks.

When the burn is terminated and the oxygen breaks up into a zillion blobs and starts floating everywhere, the oxygen under the screen umbrella cannot escape. Surface tension prevents it from escaping through the screen holes. The oxygen is trapped under the umbrella, inside the retention reservoir.

When the engines are restarted there is oxygen right at the pipe to feed into the engine, instead of a void with random floating blobs. The engine thrust then settles the oxygen in the sump tank for normal operation.

As near as I can figure, the 56.4% ullage limit happens when the storage tank is empty, so the sump tank is only partially full. But I'm not sure.

Heat Shield

Aerobraking is used to get rid of a portion of a spacecraft's velocity without using a rocket engine and reaction mass. Or as NASA thinks of it: "For Free!" This can be used for landing, for planetary capture, for circulating spacecraft's orbit, or other purposes.

Robert Zubrin says mass of the heat shield and thermal structure will be about 15% of the total mass being braked.

The general rule is that aerobraking can kill a velocity approximately equal to the escape velocity of the planet where the aerobraking is performed (10 km/s for Venus, 11 km/s for Terra, 5 km/s for Mars, 60 km/s for Jupiter).

This will mostly be used for our purposes designing a emergency re-entry life pod, not a Solar Guard patrol ship. With a sufficiently advanced engine it is more effective just to carry more fuel, so our atomic cruiser will not need to waste mass on such a primitive device.

NASA on the other hand uses aerobraking every chance it gets, since they do not have the luxury of using atomic engines. Many of the Mars probes use aerobraking for Mars capture and to circularize their orbit. Some use their solar panels as aerobraking drage chutes in order to make a given piece of payload mass do double duty. Some of the Space Tug designs listed in the Realistic Design section economize on reaction mass by using a ballute when returning to Terra orbit.

In the movie 2010, the good ship Leonov had a one-lung propulsion system, so they needed an aerobraking ballute to slow them into Jovian orbit. If you are thinking about aerobraking, keep in mind that many worlds in the Solar System do not have atmospheres.

Power Generation

If you cannot tap your propulsion system for electrical power, you will need a separate power plant (or it's going to be real dark inside your spacecraft).

Typically the percentage of spacecraft dry mass that is power systems is 28% for NASA vessels.

Spacecraft power systems have three subsystems:

  • Power Generation/ Conversion: generating power
  • Energy Storage: storing power for future use
  • Power Management and Distribution (PMAD): routing the power to equipment that needs it

There are a couple of parameters used to rate power plant performance:

  • Alpha : (kg/kW) power plant mass in kilograms divided by kilowatts of power. So if a solar power array had an alpha of 90, and you needed 150 kilowatts of output, the array would mass 90 * 150 = 13,500 kg or 13.5 metric tons
  • Specific Power : (W/kg) watts of power divided by power plant mass in kilograms (i.e., (1 / alpha) * 1000)
  • Specific Energy : (Wh/kg) watt-hours of energy divided by power plant mass in kilograms
  • Energy Density : (Wh/m3) watt-hours of energy divided by power plant volume in cubic meters

NASA has a rather comprehensive report on various spacecraft power systems here . The executive summary states that currently available spacecraft power systems are "heavy, bulky, not efficient enough, and cannot function properly in some extreme environments."

Scroll to see rest of infographic

Energy Harvesting

Energy Harvesting or energy scavenging is a pathetic "waste-not-want-not" strategy when you are desperate to squeeze every milliwatt of power out of your system. This includes waste engine heat (gradients), warm liquids, kinetic motion, vibration, and ambient radiation. This is generally used for such things as enabling power for remote sensors in places where no electricity is readily available.

Fuel Cells

The general term is "chemical power generation", which means power generated by chemical reactions. This is most commonly seen in the form of fuel cells, though occasionally there are applications like the hydrazine-fired gas turbines that the Space Shuttle uses to hydraulically actuate thrust vector vanes.

Fuel cells basically consume hydrogen and oxygen to produce low voltage electricity and water. They are quite popular in NASA manned spacecraft designs. Each PC17C fuel-cell stack in the Shuttle Orbiter has an alpha of about 10 kg/kW, specific power 98 W/kg, have a total mass of 122 kg, have an output of 12 kW, and produces about 2.7 kilowatt-hours per kilogram of hydrogen+oxygen consumed (about 70% efficient). They also have a service life of under 5000 hours. The water output can be used in the life support system.

Different applications will require fuel cells with different optimizations. Some will need high specific power (200 to 400 W/kg), some will need long service life (greater than 10,000 hours), and others will require high efficiency (greater than 80% efficient).

Solar Thermal Power

Back in the 1950's, on artist conceptions of space stations and space craft, one would sometimes see what looked like mirrored troughs. These were "mercury boilers", a crude method of harnessing solar energy in the days before photovoltaics. The troughs had a parabolic cross section and focused the sunlight on tubes that heated streams of mercury. The hot mercury was then used in turbines to generate electricity.

These gradually vanished from artist conceptions and were replaced by nuclear reactors. Generally in the form of a long framework boom sticking out of the hub, with a radiation shadow shield big enough to shadown the wheel.

The technical name is "solar dynamic power", where mirrors concentrate sunlight on a boiler. "Solar static power" is Photovoltaic solar cells.

Such systems are generally useful for power needs between 20 kW and 100 kW. Below 20 kW a solar cell panel is better. Above 100 kW a nuclear fission reactor is better.

They typically have an alpha of 250 to 170, a collector size of 130 to 150 watts per square meter at Terra orbit (i.e., about 11% efficient), and a radiator size of 140 to 200 watts per square meter.

He wasn't surprised when he was assigned to the job of helping paint the solar mirror. This was a big trough that was to run all around the top of the station, set to face the Sun. It was curved to focus the rays of the Sun on a blackened pipe that ran down its center. In the pipe, mercury would be heated into a gas, at a temperature of thirteen hundred degrees Fahrenheit. This would drive a highly efficient "steam" turbine, which would drive a generator for the needed power. When all its energy was used, the mercury would be returned to the outside, to cool in the shadow of the mirror, condensing back to a liquid before re-use.

It was valuable work, and the station badly needed a good supply of power. But painting the mirror was done with liquid sodium. It was a silvery metal that melted easily at a low temperature. On Earth, it was so violently corrosive that it could snatch oxygen out of water. But in a vacuum, it made an excellent reflective paint. The only trouble was that it had to be handled with extreme caution.

It was nasty work. A drop on the plasticized fabric of the space suits would burn a hole through them almost at once. Or a few drops left carelessly on the special gloves they wore for the job could explode violently if carried into the hut, to spread damage and dangerous wounds everywhere nearby.

Jim worked on cautiously, blending his speed with safety in a hard-earned lesson. But the first hour after the new man came out was enough to drive his nerves to the ragged edge. At first, the man began by painting the blackened pipe inside the trough.

Jim explained patiently that the pipe was blackened to absorb heat, and that the silver coating ruined it. He had to go back and construct a seat over the trough on which he could sit without touching the sodium, and then had to remove the metal chemically.

Finally, he gave up. The man was one of those whose intelligence was fine, but who never used it except for purely theoretical problems. He was either so bemused by space or so wrapped up in some inner excitement over being there that he didn't think—he followed orders blindly.

"All right," he said finally. "Go back to Dan and tell him Terrence and I can do it alone. Put your paint in the shop, and mark it dangerous. I'll clean up when I come in."

He watched the man leave, and turned to the boy who had been working with him.

Then suddenly Terrence dropped his brush into the sodium and pointed, his mouth open and working silently.

Jim swung about to see what was causing it, and his own mouth jerked open soundlessly.

The roof of the hut ahead of them was glowing hotly, and as they watched, it suddenly began crumbling away, while a great gout of flame rushed out as the air escaped. Oxygen and heat were fatal to the magnesium alloy out of which the plates were made.

The fire had been coming from the second air lock, installed when the hut was extended. The old one still worked, and men were inside the hut, laboring in space suits. An automatic door had snapped shut between the two sections at the first break in the airtight outer sheathing. But there were still men inside where the flames were, and they were being dragged out of a small emergency lock between the two sections.

One of them yanked off his helmet to cough harshly. His face was burned, but he seemed unaware of it. "Kid came through the lock with a can of something. He tripped, spilled it all over—and then it exploded. We tried to stop it, but it got away. The kid—"

He shuddered, and Jim found that his own body was suddenly weak and shaky. The third man must have done it. He'd taken the orders too literally—he'd gone to report to Dan first, before putting away the sodium. A solid hour's lecture on the dangers of the stuff had meant nothing to him.

From Step to the Stars by Lester Del Rey (1954)

Solar Photovoltaic Power

At Terra's distance to the sun, solar energy is about 1366 watts per square meter. This energy can be converted into electricity by photovoltaics. Of course the power density goes down the farther from the Sun the power array is located.

The technical name is "solar static power", where photovoltaic solar cells convert sunlight into electricity. "Solar dynamic power" is where mirrors concentrate sunlight on a boiler.

Solar power arrays have an alpha ranging from 100 to 1.4 kg/kW. Body-mounted rigid panels an alpha of 16 kg/kW while flexible deployable arrays have an alpha of 10 kg/kW. Most NASA ships use multi-junction solar cells which have an efficiency of 29%, but a few used silicon cells with an efficiency of 15%. Most NASA arrays output from 0.5 to 30 kW.

Some researchers (Dhere, Ghongadi, Pandit, Jahagirdar, Scheiman) have claimed to have achieved 1.4 kg/kW in the lab by using Culn1-×Ga×S2 thin films on titanium foil. Rob Davidoff is of the opinion that a practical design with rigging and everything will be closer to 4 kg/kW, but that is still almost three times better than conventional solar arrays.

In 2015 researchers at Georgia Institute of Technology demonstrated a photovoltaic cell using an optical rectenna. They estimate that such rectennas could have a power conversion efficiency of up to 40% and a lower cost than silicon cells. No word on the alpha, though.

The International Space Station uses 14.5% efficient large-area silicon cells. Each of the Solar Array Wings are 34 m (112 ft) long by 12 m (39 ft) wide, and are capable of generating nearly 32.8 kW of DC power. 19% efficiency is available with gallium arsenide (GaAs) cells, and efficiencies as high as 30% have been demonstrated in the laboratory.

To power a ion drive or other electric propulsion system with solar cells is going to require an array capable of high voltage (300 to 1000 volts), high power (greater than 100 kW), and a low alpha (2 to 1 kg/kW).

Obviously the array works best when oriented face-on to the sun, and unshadowed. As the angle increases the available power decreases in proportion to the cosine of the angle (e.g., if the array was 75° away from face-on, its power output would be Cos(75°) = 0.2588 or 26% of maximum). Solar cells also gradually degrade due to radiation exposure (say, from 8% to 17% power loss over a five year period if the panel is inhabiting the deadly Van Allen radiation belt, much less if it is in free space).

Typically solar power arrays are used to charge batteries (so you have power when in the shadow of a planet). You should have an array output of 20% higher voltage than the battery voltage or the batteries will not reliably charge up. Sometimes the array is used instead to run a regenerative fuel cell.

Solar Power
PlanetSol Dist
☿ Mercury0.3876.6779,121
⊕ Terra1.0001.0001,366
⚶ Vesta2.3620.179245
⚵ Juno2.6700.140192
⚳ Ceres2.7680.131178
⚴ Pallas2.7720.130178
♃ Jupiter5.2000.03751
♄ Saturn9.5800.01115
♅ Uranus19.2000.0034
♆ Neptune30.0500.0012


Like all non-coherent light, solar energy is subject to the inverse square law. If you double the distance to the light source, the intensity drops by 1/4.

Translation: if you travel farther from the sun than Terra orbit, the solar array will produce less electricity. Contrawise if you travel closer to the sun the array will produce more electricity. This is why some science fiction novels have huge solar energy farms on Mercury; to produce commercial quantities of antimatter, beamed power propulsion networks, and other power-hungry operations.

As a general rule:

Es = 1366 * (1 / Ds2)


  • Es = available solar energy (watts per square meter)
  • Ds = distance from the Sun (astronomical units)
  • 1366 = Solar Constant (watts per square meter)

Remember that you divide distance in meters by 1.496e11 in order to obtain astronomical units. Divide distance in kilometers by 1.496e8 to obtain astronomical units


What is the available solar energy at the orbit of Mars?

Mars orbits the sun at a distance of 2.28e11 meters. That is 2.28e11 / 1.49e11 = 1.52 astronomical units. So the available solar energy is:

  • Es = 1366 * (1 / Ds2)
  • Es = 1366 * (1 / 1.522)
  • Es = 1366 * (1 / 2.31)
  • Es = 1366 * 0.423
  • Es = 591 watts per square meter

This means that the available solar energy around Saturn is a pitiful 15 W/m2. That's available energy, if you tried harvesting it with the 29% efficient ISS solar cell arrays you will be lucky to get 4.4 W/m2. Which is why the Cassini probe used RTGs.

Special high efficiency cells are needed in order to harvest worthwhile amounts of solar energy in low intensity/low temperature conditions (LILT). Which is defined as the solar array located at 3 AU from Sol or farther (i.e., about 150 watts per square meter or less, one-ninth the energy available at Terra's orbit).

Equation Derivation

If you are curious where the "1,366 W/m2" Solar Constant value in the equation came from (for instance, if you want to calculated it for another star), read on. Otherwise skip this section.

It starts with the Stefan–Boltzmann law:

j = σ * T4


j total energy radiated from black body (W/m2)
σ Stefan–Boltzmann constant (5.670367×10−8 W·m−2·K−4)
T thermodynamic temperature (K)

The Sun's thermodynamic temperature is 5,778 K (effective temperature in the photosphere). Doing the math reveals that j = 63,200,617 W/m2.

To calcuate what this is at Terra's orbit (1 Astronomical Unit) we use the Inverse-Square Law. For this purpose the equation is:

P1au = (Dss2 / Dau2) * j


P1au solar power at 1 AU or solar constant (W/m2)
Dss distance from center of sun to sun's surface, the sun's radius (AU)
Dau solar constant distance (AU) = 1 AU for all stars, by definition
j total energy irradiated, from first equation (W/m2)
x2 square of x, that is x * x

The Sun's radius is 696,342 km. Dividing by 1.496e8 tells us the Sun's radius is 0.00465 AU (because the equation wants both distances in AU). Plugging it all into the equation:

P1au = (Dss2 / Dau2) * j
P1au = (0.004652 / 12) * 63,200,617
P1au = (0.0000216225 / 1) * 63,200,617
P1au = 0.0000216225 * 63,200,617
P1au = 1,369 W/m2

which is close enough for government work to 1,366 W/m2.

To calculate this for other stars you will need that star's thermodynamic temperature and radius. If you do not want to do the math, I made a quick table for you.

  1. Refer to the Star Table
  2. Look up the star's Spectral Class (the Sun is a G2 star)
  3. For thermodynamic temperature T, use value for Teeff (G2 is 5770)
  4. For Dss take the value for R (G2 is 1.0) and multiply it by 0.00465 to get star's radius in AU
  5. Calculate P1au using the two equations above

The solar array drop off equation for that star will then be:

Es = P1au * (1 / Ds2)

For example, the star Sirus A is spectral class A0. From the table Teeff is 10,000 K, use that for T. From the table R is 2.7, times 0.00465 means Dss is 0.01257.

Doing the math, J = 567,036,700 and P1au = 89,561. So for Sirus the solar array drop off equation is

Es = 89,561 * (1 / Ds2)

This means that a spacecraft with a solar array orbiting 5 astronomical units from Sirius (orbital radius of Jupiter) could harvest 3,582 watts per square meter, or about 2.6 times as much as it could get in the solar system at Terra orbit.

A more exotic variant on solar cells is the beamed power concept. This is where the spacecraft has a solar cell array, but back at home in orbit around Terra (or Mercury) is a a huge power plant and a huge laser. The laser is fired at the solar cell array, thus energizing it. It is essentially an astronomically long electrical extension cord constructed of laser light. It shares the low mass advantage of a solar powered array. It has an advantage over solar power that the energy per square meter of array can be much larger.

It has the disadvantage that the spacecraft is utterly at the mercy of whoever is currently running the laser battery. It has the further disadvantage of being frowned upon by the military, since they take a dim view of weapons-grade lasers in civilian hands. Unless the military owned the power lasers in the first place.

Radioisotope Thermoelectric Generators

Radioisotope thermoelectric generators (RTG) are slugs of radioisotopes (usually plutonium-238 in the form of plutonium oxide) that heat up due to nuclear decay, and surrounded by thermocouples to turn the heat gradient into electricity (it does NOT turn the heat into electricity, that's why the RTG has heat radiator fins on it.).

There are engineering reasons that currently make it impractical to design an individual RTG that produces more than one kilowatt. However nothing is stopping you from using several RTGs in your power room. Engineers are trying to figure out how to construct a ten kilowatt RTG.

Current NASA RTGs have a useful lifespan of over 30 years.

Currently RTGs have an alpha of about 200 kg/kW (though there is a design on the drawing board that should get about 100 kg/kW). Efficiency is about 6%. The near term goal is to develop an RTG with an alpha of 100 to 60 kg/kW and an efficiency of 15 to 20%.

An RTG based on a Stirling cycle instead of thermionics might be able to reach an efficiency of 35%. Since they would need less Pu-238 for the same electrical output, a Sterling RTG would have only 0.66 the mass of an equivalent thermocouples RTG. However NASA is skittish about Sterling RTGs since unlike conventional ones, Sterlings have moving parts. Which are yet another possible point of failure on prolonged space missions.

Nuclear weapons-grade plutonium-239 cannot be used in RTGs. Non-fissionable plutonium-238 has a half life of 85 years, i.e., the power output will drop to one half after 85 years. To calculate power decay:

P1 = P0 * 0.9919^Y


  • P1 = current power output (watts)
  • P0 = power output when RTG was constructed (watts)
  • Y = years since RTG was constructed.

If a new RTG outputs 470 watts, in 23 years it will output 470 x 0.9919^23 = 470 x 0.83 = 390 watts

Wolfgang Weisselberg points out that this equation just measures the drop in the power output of the slug of plutonium. In the real world, the thermocouples will deteriorate under the constant radioactive bombardment, which will reduce the actual electrical power output even further. Looking at the RTGs on NASA's Voyager space probe, it appears that the thermocouples deteriorate at roughly the same rate as the plutonium.

Plutonium-238 has a specific power of 0.56 watts/gm or 560 watts per kilogram, so in theory all you would need is 470 / 560 = 0.84 kilograms. Alas, the thermoelectric generator which converts the thermal energy to electric energy has an efficiency of only 6%. If the thermoelectric efficiency is 6%, the plutonium RTG has an effective specific power of 560 x 0.06 = 30 watts per kilogram 238Pu (0.033 kilogram 238Pu per watt or 33 kgP/kW). This means you will need an entire 15.5 kilos of plutonium to produce 470 watts.

This is why a Sterling-based RTG with an efficience of 35% is so attractive.

Many RTG fuels would require less than 25 mm of lead shielding to control unwanted radiation. Americium-241 would need about 18 mm worth of lead shielding. And Plutonium-238 needs less than 2.5 mm, and in many cases no shielding is needed as the casing itself is adequate. Plutonium is the radioisotope of choice but it is hard to come by (due to nuclear proliferation fears). Americium is more readily available but lower performance.

At the time of this writing (2014) NASA has a severe Pu-238 problem. NASA only has about 16 kilograms left, you need about 4 kg per RTG, and nobody is making any more. They were purchasing it from the Russian Mayak nuclear industrial complex for $45,000 per ounce, but in 2009 the Russians refused to sell any more.

NASA is "rattled" because they need the Pu-238 for many upcoming missions, they do not have enough on had, and Congressional funding for creating Pu-238 manufacturing have been predictably sporadic and unreliable.

The European Space Agency (ESA) has no access to Pu-238 or RTGs at all. This is why their Philae space probe failed when it could not get solar power. The ESA is accepting the lesser of two evils and is investing in the design and construction of Americium-241 RTGs. Am-241 is expensive, but at least it is available.

Nuclear Fission Reactors

Los Alamos reactor
Fuel region157 kg
Reflector154 kg
Heat pipes117 kg
Reactor control33 kg
Other support32 kg
Total Reactor mass493 kg

For a great in-depth analysis of nuclear power for space applications, I refer you to Andrew Presby's engineer degree thesis: Thermophotovoltaic Energy Conversion in Space Nuclear Reactor Power Systems . There is a much older document with some interesting designs here .

As far as the nuclear fuel required, the amount is incredibly tiny. Which in this case means burning a microscopic 0.01 grams of nuclear fuel per second to produce a whopping 1000 megawatts! That's the theoretical maximum of course, you can find more details here.

Nuclear fission reactors are about an alpha of 18 kg/kW. However, Los Alamos labs had an amazing one megawatt Heat Pipe reactor that was only 493 kg (alpha of 0.493 kg/kW):

Fission reactors are attractive since they have an incredibly high fuel density, they don't care how far you are from the Sun nor if it is obscured, and they have power output that makes an RTG look like a stale flashlight battery. They are not commonly used by NASA due to the hysterical reaction of US citizens when they hear the "N" word. Off the top of my head the only nuclear powered NASA probe currently in operation is the Curiosity Mars Rover; and that is an RTG, not an actual nuclear reactor.

For a space probe a reactor in the 0.5 to 5 kW power range would be a useful size, 10 to 100 kW is good for surface and robotic missions, and megawatt size is needed for nuclear electric propulsion.

Here is a commentary on figuring the mass of the reactor of a nuclear thermal rocket by somebody who goes by the handle Tremolo:

Now, onto a more practical means for generation 1 MW of power using a Plutonium fission reaction.

To calculate the mass required to obtain a certain power level, we have to know the neutron flux and the fission cross-section. Let's assume the flux is 1E14 neutron/cm2/sec, the cross section for fast fission of Pu-239 is about 2 barns (2E-24 cm2), the energy release per fission is 204 MeV, and the Pu-239 number density is 4.939E22 atoms/cm3. Then the power is

P = flux * number density * cross section * Mev per fission * 1.602E-13 Watt/MeV

P = 1E14 * 4.939E22 * 2E-24 * 204 * 1.602E-13 = 323 W/cm3

So, for 1 MW, we need 1E6/323 = 3100 cm3. Given a density of 19.6 gm/cm3, this is 19.6*3100 = 60,760 gm or 60.76 kg.

The next question to ask is: how long do you want to sustain this reaction? In other words, what is the total energy output?

For example, a Watt is one Joule per second. So, to sustain a 1 MW reaction for 1 year, the total energy is 1E6 J/s * 3.15E7 s/year = 3.15E13 J.

For Pu-239, we have 204 Mev per fission and we have 6.023E23./239 = 2.52E21 atoms/gm. So, the energy release per gram is 2.52E21 * 204 Mev/fission * 1.602E-13 J/Mev = 8.24E10 J/gm.

Therefore, to sustain 1 MW for 1 year, we will use 3.15E13 J / 8.24E10 J/gm = 382 gm of Pu-239 or 0.382 kg. This is only a small fraction of the total 60.76 kg needed for the fission reaction.

Finally, this is thermal energy. Our current light water reactors have about a 35% efficiency for conversion to electric power. So, you can take these numbers and essentially multiply by 3 to get a rough answer for the total Pu-239 needed: 3 x 60.76 = 182 kg. Rounding up, you would need roughly 200 kg for a long term sustained 1 MW fission reaction with a 35% conversion efficieny.

These calculations assume quite a bit and I wouldn't use these numbers to design a real reactor, but they should give you a ballpark idea of the masses involved.


New reactors that have never been activated are not particularly radioactive. Of course, once they are turned on, they are intensely radioactive while generating electricity. And after they are turned off, there is some residual radiation due to neutron activation of the reactor structure.

How much deadly radiation does an operating reactor spew out? That is complicated, but Anthony Jackson has a quick-and-dirty first order approximation:

r = (0.5*kW) / (d2)


  • r = radiation dose (Sieverts per second)
  • kW = power production of the reactor core, which will be greater than the power output of the reactor due to reactor inefficiency (kilowatts)
  • d = distance from the reactor (meters)

This equation assumes that a 1 kW reactor puts out an additional 1.26 kW in penetrating radiation (mostly neutrons) with an average penetration (1/e) of 20 g/cm2.

As a side note, in 1950's era SF novels, nuclear fission reactors are commonly referred to as "atomic piles." This is because the very first reactor ever made was basically a precision assembled brick-by-brick pile of graphite blocks, uranium fuel elements, and cadmium control rods.


Space Nuclear Power Program

This program aims to develop high-power, safe, reliable nuclear energy sources for manned and deep-space missions. Nuclear electric propulsion will open a new frontier of exploration in the outer solar system and allow manned missions to Mars and other places with difficult solar power problems.

A starting point might be the direct gas reactor studied for the Prometheus project, a 1MWt/200kWe reactor at 40-50kg/kW (7.5-11t). Another might be the heatpipe reactor SAFE-400, a 400kWt/100kWe reactor at unknown specific mass. Sodium-cooled designs in the 70kg/kW range are also possible. (Masses include radiators, conversion and power conditioning.)

The SAFE-30 project demonstrated simple, affordable ground testing of non-nuclear components. The Prometheus project demonstrated productive cooperation with Naval Reactors and related organizations to tap their nuclear technology expertise. Joining these approaches will allow the project to proceed immediately into materials testing and design optimization. The most urgently needed component is an experimental fast reactor for materials testing. Also critical will be a design process that focuses on modular power units so the same basic design can be used for a wide range of missions, presumably in the 50-100kWe range.

Costs are not straightforward to estimate. One baseline figure is the $4.2 billion estimated to develop the Prometheus reactor system. Let’s assume a 50% increase on that figure and use $6.3 billion for the development program; further assume hardware costs of $5000 per watt. Three demonstration units will be built: one for flight test (possibly on a later carrier flight), one for a NEP asteroid capture and one as a base power supply for a manned mission. The goal is a 50kWe power unit massing 2,000kg or better (40kg/kW) with at least 20-year useful life. Individual units could power NEP asteroid retrieval tugs or small ISRU operations; sets of four could power manned bases or deep-space probes. A second phase using knowledge gained from the first generation reactor program would aim to build power units of 1MWe and 10t mass range (~10kg/kW) for use on deep-space and interstellar probes, permanent bases and orbital manufacturing facilities. All future NEP missions would be able to use a proven, existing design and avoid developmental uncertainties.

Estimated costs:

$6,300m development program
$750m flight hardware
$2,115m margin
$9,165m total cost ($611m per year)

Alternate scenario: A fast-spectrum reactor is made available by another country or organization for materials testing. Majority of the design, testing and construction is outsourced to Naval Reactors and experienced contractors. Additional funding is provided by ESA and allied space agencies in return for access to flight hardware. Development program costs cut in half and a fourth power unit is built for ESA use. New costs:

$3,150m development program
$750m flight hardware
$1,170m margin
$5,070m total cost ($338m per year)

 This is a subject that's been stewing for a while now. I often see debates in comment sections over whether or not nuclear electric power is feasible in space. Only rarely do those arguing hold the same assumptions about what nuclear power actually means. As a result, these debates rarely convince anyone of anything beyond the stubborn natures of their opponents.

 The goal of this post is to briefly cover the range of commercial, military and scientific nuclear power systems ranging from a few kilowatts to over a gigawatt. I will follow up the (hopefully) useful background information in a later post with some fanciful projections and my usual call for unlikely investments in space.

Very briefly:
Nuclear energy is produced by the fission (splitting) of certain heavy atoms. This fission produces radiation which becomes heat which is then turned into electricity. The leftover heat and spent nuclear fuel must be dealt with. Shielding must be provided.


 I won't get too deep into this subject, but there are several types of radiation. All of these types create challenges for material designs, since most materials become brittle with exposure to radiation. (Would you like to know more?)

 - Neutrons are nuclear particles emitted during fission; a certain amount of neutron radiation is needed to start up most nuclear reactors. Neutrons can be either fast (high energy) or thermal (fast neutrons that have been slowed by smashing into a moderator). Neutrons are a form of penetrating radiation; they are a neutral particle so electrical interactions have no effect, which means they can penetrate deep into many materials. Neutrons can also 'activate' other materials; once a neutron has been slowed down by many collisions with atoms, it eventually gets slow enough to be captured. This neutron capture process can form radioactive isotopes of common materials like iron or nitrogen. The best shielding for neutrons is either a lot of hydrogen (usually as water or polyethylene) or layers of neutron reflectors (lead, bismuth, beryllium; see below). It's important to note that the neutron environment inside a reactor must be carefully controlled for efficient operation, and there is definitely a lower limit as well as an upper limit for workable designs.
 - Gamma rays are very high energy photons (electromagnetic energy) produced either directly during fission, indirectly after a positron (anti-electron) is released and then annihilated with an electron, or indirectly by a beta particle colliding and emitting bremsstrahlung. Gamma is undesirable in a reactor because it is penetrating, very harmful and can be activating. Gamma rays can trigger the fission of deuterium, for example, causing the release of a moderate-energy neutron. The best shielding for gamma is a heavy metal like tungsten, but often a conductive liner (steel) and a bulk absorber (very thick concrete) are used.
 - Other particles (protons, alpha particles and heavier fission fragments) have different typical energy levels but are largely the same as far as a reactor is concerned. They are typically charged, can be slowed or stopped efficiently with metals and eventually become troublesome atoms trapped inside the fuel or coolant. Higher-speed fragments will also emit bremsstrahlung as they slow down, so essentially all nuclear reactors produce some level of gamma radiation.


 The simplest fission fuel is an unstable isotope that spontaneously decays. Plutonium-238 is probably the most common example; this is used in RTG (radioisotope thermoelectric generator) units and radioactive heater units on deep space probes. Strontium-90 is another example, widely used in the Soviet Union in space and on Earth as a reliable power source for remote outposts like lighthouses. Some additional possibilities are Polonium-210 (powerful, dangerous, short life) and Americium-241 (long life, relatively high penetrating radiation output). These decay fuels are usually used as a simple source of heat, either maintaining operating temperature for some other device or powering a thermoelectric generator. The ideal unstable fuel would be something that decays only into alpha particles and stable products, producing no penetrating or activating radiation while having a decay rate high enough to be reasonably energy-dense yet low enough to operate for a few decades. No such material is known.

 Next is fissile material. A fissile isotope is one that can capture a low-energy neutron and then split. The four main examples are uranium-235 (naturally occurring), uranium-233 (bred from thorium-232), plutonium-239 (bred from uranium-238) and plutonium-241 (bred from plutonium-239 by way of Pu-240). Fissile material is useful for making nuclear weapons, so the production and use of these isotopes is very tightly controlled. Inefficient early reactors couldn't use natural uranium because the fissile content was too low; the U-235 had to be separated (enriched) to produce a fuel that would work properly. The same technology is used to make highly-enriched material for weapons, so again enrichment technology is tightly controlled. More modern reactor designs are more neutron-efficient, so they can use fuel that is less enriched or not enriched at all. Note that highly-enriched fissile material is very dangerous to handle or transport; too much of it in one place or accidentally exposed to neutron flux could lead to a chain reaction, a sudden spike in radioactivity and heat.

 Last is fertile material. A fertile isotope is one that can capture a neutron and convert into a fissile isotope, which can then be split with another neutron. Examples are uranium-234 (natural, makes U-235), uranium-238 (natural, makes U-239), thorium-232 (natural, makes U-233), plutonium-238 (artificial, makes Pu-239) and plutonium-240 (artificial, makes Pu-241). Fertile materials are relatively stable; they are not particularly radioactive nor will they do anything dangerous if you put a lot of it in one place. Most of them are flammable metals, but that is a chemical hazard rather than a nuclear hazard; burning U-238 is no more dangerous than burning magnesium (though the results are a bit more toxic). Fertile materials (including natural uranium) are far easier to transport safely than fissile or unstable materials.

Fuel Cycles

 The fuel by itself is only part of the story. The full fuel cycle is important to consider. Earth-based commercial power reactors can rely on an extensive infrastructure of mining, refining, enrichment, fabrication, reprocessing and disposal. Space-based reactors will have none of those advantages.

 Most commercial reactors and some military reactors are thermal, meaning their fast neutrons are moderated down to an energy level that allows for efficient capture in fissile fuel. Most such reactors require enriched fuel, which means fuel elements would be shipped from Earth until nuclear materials processing infrastructure is established in space. This is politically, economically and environmentally difficult, so Earth-style thermal reactors are not likely to be used in space for a long time if ever. One possible exception is CANDU, a heavy water moderated thermal reactor that can burn natural uranium (and a lot of other radioactives) as fuel. Interestingly, ice on Mars is significantly richer in heavy water than on Earth thanks to atmospheric losses over the eons; this might be a reasonable medium-term approach, particularly since the design does not require massive pressure vessels.

 Many research and medical reactors and some military reactors are fast, meaning their neutrons are used as they are produced. Fast reactors are often called breeder reactors, because they turn fertile material into fissile material which is then split for energy. An initial 'spark plug' of fissile material is used to generate enough neutrons to get the reactor going, then the majority of the fuel is natural uranium, natural thorium or some other fertile material. The earliest breeder reactors were used to generate fissile plutonium for the production of nuclear weapons, but current designs using thorium are specifically intended to prevent any application to weapons (proliferation-safe). Small-scale research and medical reactors are used to irradiate materials to make useful isotopes for medical imaging, cancer radiation therapy and RTG power cores. Thorium-based reactors are particularly interesting for space colonization since they could be fueled using rudimentary refining techniques and produce little waste.

Moderators, Coolants, Poisons and Reflectors

 The neutron environment inside a reactor is critically important to safe and efficient operation. Four types of materials are present in most reactors and all of them affect how neutrons behave. Many materials have more than one property from this group.

 A moderator is some material that can absorb energy from neutrons without stopping them entirely. A coolant is something that can carry heat efficiently and hopefully is not too corrosive or degraded by radiation. By far the most common material in both cases is plain water thanks to its high hydrogen content, excellent heat capacity and reasonable thermal conductivity. Commercial power reactors are almost exclusively thermal, either pressurized water or boiling water types, which use purified light water to moderate neutrons and to carry heat out of the core. Care must be taken that the design is passively safe; that is, if the coolant were to boil suddenly then the reactor should naturally reduce its power output without intervention. For an example of passive safety, check out TRIGA (training, research, isotopes, General Atomics) reactors; operating safely since 1958 these are the only reactors licensed for unattended operation.

 The two other moderators in common use are heavy water (water made of oxygen and deuterium) and graphite (pure carbon). A third used in a handful of experimental and military reactors is lithium-7 (with or without beryllium), typically as part of a molten salt.
 The main heavy water reactor design is CANDU, which uses it as both moderator and coolant. Derivative designs use separate light and heavy water systems, with the heavy water providing mostly moderation and the light water providing mostly cooling. Heavy water is used because the hydrogen already has an extra neutron and is much less likely to capture another one. It does happen, so heavy water reactors produce small amounts of tritium.
 Graphite always uses a separate coolant since it is a solid. Graphite was used in the first reactor (the Chicago pile) and in many others since then due to its stability, mechanical strength, incredible temperature tolerance and ready availability. As a solid, graphite is susceptible to lattice defects called Wigner energy; this led to the Windscale fire before it was understood, though most modern reactors operate above the annealing temperature of carbon so this is not a concern.
 Beryllium is a suitable moderator if you only look at physics. Unfortunately it's expensive and extremely toxic, so it is not normally used on its own. In a mix with lithium-7 and fluorine it forms the coolant/moderator FLiBe used in molten salt reactors.

 Fast reactors need to have as little moderation as possible (or at least a predictable and controllable amount) inside the core. That means they need to use coolants that are poor moderators or are neutron-transparent. Common materials are sodium and lead (yes, lead; it's great at absorbing gamma radiation but it tends to reflect neutrons). Some molten salt reactors are also fast reactors and may use zirconium and sodium fluorides instead of beryllium and lithium fluorides in the salt mix. It's worth noting that some graphite-moderated reactors are cooled with molten lead or sodium, since using a coolant that is a poor moderator means the reactor's behavior is more predictable during transient problems with coolant flow.
 Carbon dioxide has been used as a coolant (with moderating properties) in the past, and may be used again as a supercritical fluid. This requires fairly high pressures, but learning how to handle supercritical CO2 would have useful applications for cooling or refrigeration elsewhere in space.
 Helium has also been used as a coolant and is proposed to be used in some very high temperature reactors as both the coolant and the working fluid for the turbine. Because it resists activation, if a reactor core uses fuel elements that trap their own fission products then the helium can pass directly through the core and into the generator turbine with no intermediate heat exchangers; this requires very high temperature turbine materials but leads to superior efficiency and compact, simple design.
 Zirconium is nearly transparent to neutrons. Many fuel assemblies use Zircalloy, an alloy that is at least 95% zirconium, to allow fast neutrons to escape the fuel pins and to allow moderated neutrons back into the fuel to trigger more fission. A common fuel is uranium zirconium hydride, with zirconium alloyed for structural strength and hydrogen adsorbed for inherent moderation.

 A poison is some material that absorbs neutrons very efficiently. Examples include lithium-6, boron, hafnium, xenon-135 and gadolinium. These are used in control rods and safety systems or are produced naturally by nuclear reactions within the core. Over time, neutron poisons build up in the fuel; the dynamics of this are complex but neutron poisons are the main reason why uranium fuels only burn about 2% of their potential in one pass through a reactor. The poison byproducts have to be removed for the fuel to become usable again. Xenon is the most important of these over short timescales.
 Hafnium, boron and gadolinium are common materials for control rods. These devices allow operators to precisely control how many neutrons are flying around at a given time inside the core and can also be used as an emergency shutdown device. Control rods may be suspended above the core by electromagnets; during a loss of electrical power the rods will naturally fall into the core and stop primary activity. Soluble boron salts are used as an emergency shutdown tool in water-moderated reactors; the salt is injected into the moderator or coolant loop, causing an immediate and dramatic reduction in neutron flux and stopping the reactor's primary activity. Radioactive byproducts will still produce significant heat and radiation for hours to days, so additional safety features like auxiliary cooling are required.

 A reflector is a material that reflects (elastically scatters) neutrons. Primary examples are beryllium, graphite, steel, lead and bismuth. This is another reason why graphite was used in early reactors: a layer of solid graphite blocks around the outside of the pile reflected neutrons back into the core, reducing the required size of the core and reducing the required neutron shielding.
 Many reactor designs intended for use in space rely on controllable reflectors rather than controllable poisons; the reactor core would be safe (subcritical) by design, only able to operate when neutron reflectors were properly placed. That allows a reactor to be launched before activation, meaning the potential radioactive release during a launch accident would be minimized.
Some other designs use reflectors to boost reactivity near the end of life for a given batch of fuel, or otherwise as an alternative to poisons for control. An example is the SSTAR design, which would use a movable reflector to move the active region of the reactor through a fuel load over the course of 30 years rather than refueling every ~18 months. If the reflector were to fail then the reactor's output would taper off to nearly nothing over a few days. By relying on reflectors rather than poisons, the reactor requires a lower level of neutron flux to operate and can use less efficient (less or not enriched) fuels.

Turning heat into electricity

 Once you have a steady supply of heat, you have to put it to use somehow. The laws of physics are singularly unforgiving about energy conversion. For every useful unit of electricity produced you will have to deal with two to five units of waste heat in any practical design. Less efficient options are always available.
 In space we don't have access to free-flowing rivers or oceans of water to use as coolant; without conduction or convection we can rely only on radiation. Thermal radiators are significantly more efficient at high temperature, so the higher our core reactor temperature the better for a free-flying spacecraft. (Radiative output scales as the fourth power of temperature, so a small increase in temperature causes a very large increase in radiator output.) The temperature limit for a reactor is usually based on either the primary coolant or the fuel material, around 900-1000 °C for zircalloy cladding and possibly higher for ceramic or carbide fuel elements. Molten salt or gas-cooled reactors could go higher, while water-cooled reactors are a fair bit lower. (Water-cooled reactors use water at high pressures, so the boiling point of the coolant is typically several hundred °C.) I won't get into the physics and mechanics of radiators here other than to say they are similar to solar panels in terms of areal density, pointing and deployment. The size of a radiator system depends very strongly on the temperature of the coolant and whether there is a large hot object (like Earth) nearby.
 For a surface base with access to a large thermal mass (dirt, ice, etc.) there may be the option of process heat. Some of the waste heat from the reactor can be used to do useful work like melting ice, heating greenhouses or powering thermochemical reactions like the sulfur-iodine process for producing hydrogen. From the perspective of the electrical generation system this is still waste energy, but these uses increase the overall efficiency of the system. This kind of cogeneration greatly increases the required radiator area in free space, so although it seems counterintuitive it may not be mass-efficient to use waste heat for chemical processes on an orbital station. Rather, it may actually take less mass to produce electricity (at 20-30% efficiency, but with high-temp radiators) and use it directly in electrochemical processes vs. thermochemical processes. Each individual mission / craft / architecture is unique and may come down on either side of the line.

 So, with a source of heat (reactor coolant loop) and a sink of heat (radiator coolant loop) we can put a heat engine between the two and extract useful energy. The most basic approach is to use the thermoelectric effect (like a Peltier cooler), directly converting heat into an electric current. These devices typically have no moving parts and are highly reliable, but are poorly scalable and only modestly efficient. RTGs use these, as have some flown reactors on Soviet satellites.
 By far the most common method on Earth is to use a steam turbine in the Rankine cycle. Heat from the reactor loop boils water into steam in a steam generator, which is passed through a turbine to rotate a shaft. The depleted steam is recondensed into water, passing low temperature waste heat into the cooling loop. This would be extremely inefficient in space as the low waste temperature would require enormous radiators.
 A promising technique is to use the Brayton cycle in a reactor with a gas coolant. The most likely of these is helium, since it is very stable and nearly impervious to neutrons. A space-optimized Brayton cycle reactor (see for example project Promethius) would circulate helium through the core and pass it directly through the turbine, with no intermediate loops or heat exchangers. This is possible only because helium does not become radioactive inside the core, but it also requires that the fuel elements contain all fission products; any fuel leak would contaminate the turbine. A cycle using steam without a condenser and boiler is also possible.
 Surface bases with abundant heatsink potential could use a Combined cycle. This is a high-temperature Brayton cycle turbine whose waste heat is still high enough to run a Rankine cycle turbine of one or two stages. The Rankine cycle exhaust heat is quite low temperature and would have to be rejected into a body of water (or some other liquid) or pumped into the ground like a reverse geothermal system. The best case would be a mixed-use system that provides electricity, industrial process heat for thermochemistry and ice melting, and life support heat for maintaining livable habitat conditions. Using an array of greenhouses as your low-temperature radiator system would be ideal. The drawbacks of a system like this are complexity, need for available heat sinks and the fact that each part of the process relies on all other parts maintaining a certain pace. If you want to have electricity while your industrial processes are not running then you need an alternate heat sink to replace those processes.

Dealing with waste

 Nuclear reactions produce radiation. Some of that radiation ends up activating parts of the reactor, which means those parts become radioactive themselves. Pumps, valves, pipes, pressure vessel walls, all of the structure in the core of a reactor will become radioactive over time. This material generally can't be reprocessed into a nonradioactive form. (It's possible but would be extremely expensive.) This is usually low to medium grade nuclear waste and the usual solution is to slag it, encase it in concrete and bury it. That probably works for surface bases on bodies with no 'weather' cycle, but it would be a no-go for active worlds like Titan / Io or icy worlds like Europa. Even then, there has to be some standardized way to indicate to future generations that there is something dangerous buried there. For craft and colonies that can't bury their waste, they would have to find some place to send it safely. This remains an unsolved problem on Earth; perhaps a waste repository and reprocessing center on the moon might some day be viable, provided shipments of waste are ever allowed to be launched.
 The fuel itself produces radioactive byproducts as a result of fission. These are mostly actinides, but there are some radioactive gases like iodine as well. On Earth we generally store fuel elements indefinitely in cooling ponds or eventually in dry casks. Fuel elements can be reprocessed, meaning the component materials are separated, byproducts are filtered out and the repurified fuel is recast into new fuel elements. The actinide wastes can be burned in certain types of reactor (usually the same sort that can burn thorium, but some fast spectrum reactors are designed for waste destruction). The old liners or shells and any equipment used in fuel processing will generally be considered high-grade nuclear waste; this is treated much like other types of waste but will be radioactive for a much longer time due to contamination with radioactive isotopes. Fuel reprocessing facilities are a proliferation concern because they allow for the extraction of weapons-grade plutonium from spent uranium fuels. Thorium cycle reactors would be politically easier because it is far more difficult to get anything of military interest out of the fuel.


 Radiation from an operational reactor is damaging to people, electronics and structures. Shielding must be provided to mitigate this damage. Earth reactors solve this problem using cheap, bulky, heavy material in abundance. Usually the reactor core is placed inside a containment building; the building is a thick stainless steel liner and several meters of concrete all around. Openings usually take sharp turns so there is no line of sight from the core to the outside world; radiation doesn't turn corners. (It does scatter, so it's still not simple.)
 Free-flying reactor designs don't have to worry about contaminating a planet full of voters during a system failure. These usually have the reactor at one end of the ship on a long truss, with a small shield plug (a shadow shield) that protects the rest of the spacecraft. Ships like these are easy to see coming if you have gamma detectors. They are great for deep space exploration, but they make bad neighbors and are difficult to handle for docking maneuvers since a small misalignment could kill everyone on the other ship.

Possible scenarios - surface base

 Let's look at the simplest case first. This is a manned surface colony with basic industry already online. Base metals (iron, nickel, aluminum) and bulk material (dirt) are available. First the coolant system is built (or installed) and tested. Next a containment pit is dug, then lined in concrete/sintered or pressed regolith/etc. Nickel-iron (simply iron from here on) blocks are piled up like bricks and welded together. A self-contained core unit is assembled on Earth and shipped in one piece, placed into the pit and connected to the radiator system. The pit is covered with iron sheets or beams with a layer of concrete/sinter/etc. then buried. The core unit is not activated until it is installed, so it is not radioactive and has no unusual handling restrictions. It would be designed to run for 20-30 years unattended, with no maintenance access possible; it would probably be limited to a few tons mass at most (~6-8t; 300-500kg fuel mass) and up to a few hundred kilowatts of electricity. New core assemblies would be shipped about every decade to maintain redundancy, more often if the colony's energy needs are growing. Cores would be in the few hundred million dollar range (plus shipping); comparable cores on Earth can be built for tens of millions but they don't need to survive a reentry accident and can be repaired on-site. Lifetime power generation (20 years, 95% availability) would be about 33 GWhr of electricity.
 The whole assembly would be several meters underground, safe to stand above while operating. A coolant failure would leave the reactor hot but safe, which means the coolant system could be rebuilt or replaced without needing to do anything to the reactor core. In the event of a serious problem like a core meltdown, any released radioactive gases would escape into space or be diffused through the (already unbreathable) atmosphere. Particles could be a bigger problem; on Mars they would be swept away in the next dust storm but on the Moon they would likely stick around for a while unless they were small enough for electrostatic scattering. Still, no crops would be contaminated.

 The next step would be an accessible reactor core that can be refueled. Fuel elements could be shipped from Earth or manufactured locally. The containment structure would not be much different, but the core could be bulkier; this would allow for things to be shipped in pieces and assembled on-site. Telerobotics would be ideal for this work, but the initial construction could be safely done in person. If the local industry is capable of building small superalloy pressure vessels then something like the CANDU approach can be used, where small tubes with fuel run through a large 'tub' of moderator+coolant at manageable pressure. Regardless, a gigawatt-sized pressure vessel is a tall order for local industry (many nations on Earth couldn't build a reactor pressure vessel today) and for in-space shipping; one way or another the approach will have to be modular and scalable. Perhaps an array of many reactor cores will feed a small number of high-power turbines. Core units will likely be in the range of a few hundred kW to about one MW each (5-25t including core coolant but not turbines).
 This modular approach would allow the colony to transition into locally-manufactured fuel elements and other parts. These might initially be reprocessed fuel from earlier cores or they could start right away with locally mined material.

 Beyond that, once the colony has the capacity to make high-performance turbines, pumps, pressure vessels, fuel assemblies, etc. then they will essentially be self-reliant.

Bimodal NTR

Nuclear Thermal Rockets are basically nuclear reactors with a thrust nozzle on the bottom. A concept called Bimodal NTR allows one to tap the reactor for power. This has other advantages. Since the reactor is running warm at a low level all the time (instead of just while thrusting) it doesn't have to be pre-heated if you have a burn coming up. This reduces thermal stress, and reduces the number of thermal cyclings the reactor will have to endure over the mission. It also allows for a quick engine start in case of emergency.

In the real world, during times of disaster, US Navy submarines have plugged their nuclear reactors into the local utility grid. This supplies emergency electricity when the municipal power plant is out. In the science fiction world, a grounded spacecraft with a bimodal NTR could provide the same service.

Dusty Plasma Fission Reactors

This is from A Half-Gigawatt Space Power System using Dusty Plasma Fission Fragment Reactor (2016)

Rodney Clarke and Robert Sheldon were working on a fission-fragment rocket engine when they noticed a useful side-benefit.

There is a remarkably efficient (84%) electrical power plant called a Magnetohydrodynamic Generator (MHD generator). They also have the virtue of being able to operate at high temperatures, and have no moving parts (which reduces the maintenance required and raises reliability). A conventional electrical power generator spins a conducting copper wire coil inside a magnetic field to create electricity. An MHD generator replaces the solid copper coil with a fast moving jet of conducting plasma.

Because many designs for fusion rocket engines and fusion power plants produce fast moving jets of plasma, MHD generators were the perfect match. Ground based power plants just sprayed the jet of fusion plasma into the MHD.

Fusion spacecraft could be bimodal. An MHD generator could be installed in the exhaust nozzle to constantly bleed off some of the thrust power in order to make electricity, this was popular with inertial confinement fusion which need to recharge huge capacitors before each fusion pulse. Alternatively the MHD generator could be installed at the opposite end of the fusion reaction chamber. The fusion plasma goes down out the exhaust nozzle for thrust, but it can be diverted upwards into an MHD generator for electrical power.

Finally getting to the point, Clarke and Sheldon realized that a fission-fragment rocket engine also produces a jet of plasma. Therefore, it too can be bimodal with the addition of an MHD generator.

Cutting to the chase, they would have a jaw-dropping specific power of 11 kWe/kg! The rough design they made had a power output of 448 megawatts and a total mass of 38,430 kg (38 metric tons).

Dusty Plasma Power Reactor
Power Output448 MW
Specific Power11 kWe/kg
Mass Schedule
U235 Fuel4.27 kg
Am242m Fuel1.25 kg
Moderator9,424 kg
Moderator Heat Radiator28,000 kg
Generator Heat Radiator1,000 kg
TOTAL38,430 kg

Fusion Reactors

A fusion reactor would produce energy from thermonuclear fusion instead of nuclear fission. Unfortunately scientist have yet to create a fusion reactor that can reach the "break-even" point (where is actually produces more energy than it consumes), so it is anybody's guess what the value for alpha will be.

The two main approaches are magnetic confinement and inertial confinement. The third method, gravitational confinement, is only found in the cores of stars and among civilizations that have mastered gravidic technology. The current wild card is the Polywell device which is a type of inertial electrostatic confinement fusion generator.

Fusion is even more efficient than fission. You need to burn 0.01 grams of fission fuel per second to generate 1000 megawatts. But among the most promising fusion fuels, they start at 0.01 grams per second, and can get as low as 0.001 grams per second. You can find more details here.

Exotic power sources

There are all sorts of exotic power sources. Some are reasonably theoretically possible, others are more fringe science. None of them currently exist, and some never will.

Beamed Power

This is where the spacecraft receives its power not from an on-board generator but instead from a laser or maser beam sent from a remote space station. This is a popular option for spacecraft using propulsion systems that require lots of electricity but have low thrusts.

For instance, an ion drive has great specific impulse and exhaust velocity, but very low thrust. If the spacecraft has to power the ion drive using a heavy nuclear reactor with lead radiation shielding, the mass of the spacecraft will increase to the point where its acceleration could be beaten by a drugged snail. But with beamed power the power generator adds zero mass to the spacecraft, since the heavy generator is on the remote station instead of onboard and laser photons weigh nothing.

The drawback includes the distance decrease in power due to diffraction, and the fact that the spacecraft is at the mercy of whoever is running the remote power station. Also maneuvers must be carefully coordinated with the remote station, or they will have difficulty keeping the beam aimed at the ship.

The other drawback is the laser beam is also a strategic weapons-grade laser. The astromilitary (if any) take a very dim view of weapons-grade laser cannon in the hands of civilians. The beamed power equipment may be under the close (armed) supervision of the Laser Guard.

Antimatter Power

Any Star Trek fan knows that the Starship Enterprise runs on antimatter. The old term is "contra-terrene", "C-T", or "Seetee". At 100% of the matter-antimatter mass converted into energy, it would seem to be the ultimate power source. The operative word in this case is "seem".

What is not as well known is that unless the situation is non-standard, antimatter is not a fuel. It is an energy transport mechanism. Unless there exist "antimatter mines", antimatter is an energy transport mechanism, not a fuel. In Star Trek, I believe they found drifts of antimatter in deep space. An antimatter source was also featured in the Sten series. In real life, astronomers haven't seen many matter-antimatter explosions. Well, they've seen a few 511 keV gamma rays (the signature of electron-positron antimatter annihilation), but they've all been from thousands of light years away and most seem to be associated with large black holes. If they are antimatter mines, they are most inconveniently located. In Jack Williamson's novels Seetee Ship and Seetee Shock there exist commercially useful chunks of antimatter in the asteroid belt. However, if this was actually true, I think astronomers would have noticed all the antimatter explosions detonating in the belt by now.

And antimatter is a very inefficient energy transport mechanism. Current particle accelerators have an abysmal 0.000002% efficiency in converting electricity into antimatter (I don't care what you saw in the movie Angels and Demons). The late Dr. Robert Forward says this is because nuclear physicist are not engineers, an engineer might manage to increase the efficiency to something approaching 0.01% (one one-hundredth of one percent). Which is still pretty lousy, it means for every megawatt of electricity you pump in to the antimatter-maker you would only obtain enough antimatter to create a mere 100 pathetic watts. The theoretical maximum is 50% due to the pesky Law of Baryon Number Conservation (which demands that when turning energy into matter, equal amounts of matter and antimatter must be created).

In Charles Pellegrino and George Zebrowski novel The Killing Star they deal with this by having the Earth government plate the entire equatorial surface of the planet Mercury with solar power arrays, generating enough energy to produce a few kilograms of antimatter a year. They do this with von Neumann machines, of course.

Of course the other major draw-back is the difficulty of carrying the blasted stuff. If it comes into contact with the matter walls of the fuel tank the resulting explosion will make a nuclear detonation seem like a wet fire-cracker. Researchers are still working on a practical method of containment. In Michael McCollum's novel Thunder Strike! antimatter is transported in torus-shaped magnetic traps, it is used to alter the orbits of asteroids ("torus" is a fancy word for "donut").

Converting the energy from antimatter annihilation into electricity is also not very easy.

The electrons and positrons mutually annihilate into gamma rays. However, since an electron has 1/1836 the mass of a proton, and since matter usually contains about 2.5 protons or other nucleons for each electron, the energy contribution from electron-positron annihilation is negligible.

For every five proton-antiproton annihilations, two neutral pions are produced and three charged pions are produced (that is, 40% neutral pions and 60% charged pions). The neutral pions almost immediately decay into gamma rays. The charged pions (with about 94% the speed of light) will travel 21 meters before decaying into muons. The muons will then travel an additional two kilometers before decaying into electrons and positrons.

This means your power converter needs a component that will transform gamma rays into electricity, and a second component that has to attempt to extract the kinetic energy out of the charged pions and convert that into electricity. The bottom line is that there is no way you are going to get 100% of the annihilation energy converted into electricity. Exactly what percentage is likely achievable is a question above my pay grade.

The main virtue of antimatter power is that it is incredibly concentrated, which drastically reduces the mass of antimatter fuel required for a given application. And mass is always a problem in spacecraft design, so any way of reducing it is welcome.

The man known as magic9mushroom drew my attention to the fact that Dr. James Bickford has identified a sort of antimatter mine where antimatter can be collected by magnetic scoops (be sure to read the comment section), but the amounts are exceedingly small. He foresees using tiny amounts of antimatter for applications such as catalyzing sub-critical nuclear reactions, instead of just using raw antimatter for fuel. His report is here.

Dr. Bickford noted that high-energy galactic cosmic rays (GCR) create antimatter via "pair production" when they impact the upper atmospheres of planets or the interstellar medium. Planets with strong magnetic fields enhance antimatter production. One would think that Jupiter would be the best at producing antimatter, but alas its field is so strong that it prevents GCR from impacting the Jovian atmosphere at all. As it turns out, the planet with the most intense antimatter belt is Earth, while the planet with the most total antimatter in their belt is Saturn (mostly due to the rings). Saturn receives almost 250 micrograms of antimatter a year from the ring system. Please note that this is a renewable resource.

Dr. Bickford calculates that the plasma magnet scoop can collect antimatter about five orders of magnitude more cost effective than generating the stuff with particle accelerators.

Keep in mind that the quantities are very small. Around Earth the described system will collect about 25 nanograms per day, and can store up to 110 nanograms. That has about the same energy content as half a fluid ounce of gasoline, which ain't much. However, such tiny amounts of antimatter can catalyze tremendous amounts of energy from sub-critical fissionable fuel, which would give you the power of nuclear fission without requiring an entire wastefully massive nuclear reactor. Alternatively, one can harness the power of nuclear fusion with Antimatter-Catalyzed Micro-Fission/Fusion or Antimatter-Initiated Microfusion. Dr. Bickford describes a mission where an unmanned probe orbits Earth long enough to gather enough antimatter to travel to Saturn. There it can gather a larger amount of antimatter, and embark on a probe mission to the outer planets.

Vacuum energy

Vacuum energy or zero-point energy is one of those pie-in-the-sky concepts that sounds too good to be true, and is based on the weirdness of quantum mechanics. The zero-point energy is the lowest energy state of any quantum mechanical system, but because quantum systems are fond of being deliberately annoying their actual energy level fluctuates above the zero-point. Vacuum energy is the zero-point energy of all the fields of space.

Naturally quite a few people wondered if there was a way to harvest all this free energy.

Currently the only suggested method was proposed by the late Dr. Robert Forward, the science fiction writer's friend (hard-SF writers would do well to pick up a copy of Forward's Indistinguishable From Magic). His paper is Extracting Electrical Energy From the Vacuum by Cohesion of Charged Foliated Conductors, and can be read here.

How much energy are we talking about? Nobody knows. Estimates based on the upper limit of the cosmological constant put it at a pathetic 10-9 joules per cubic meter (about 1/10th the energy of a single cosmic-ray photon). On the other tentacle estimates based on Lorentz covariance and with the magnitude of the Planck constant put it at a jaw-dropping 10113 joules per cubic meter (about 3 quintillion-septillion times more energy than the Big Bang). A range between 10-9 and 10113 is another way of saying "nobody knows, especially if they tell you they know".

Vacuum energy was used in All the Colors of the Vacuum by Charles Sheffield, Encounter with Tiber by Buzz Aldrin John Barnes, and The Songs of Distant Earth by Sir Arthur C. Clarke.

Arguably the Grand Unified Theory (GUT) drives and GUTships in Stephen Baxter's Xeelee novels are also a species of vacuum energy power sources.


Casimir batteries and engines

A common assumption is that the Casimir force is of little practical use; the argument is made that the only way to actually gain energy from the two plates is to allow them to come together (getting them apart again would then require more energy), and therefore it is a one-use-only tiny force in nature. In 1984 Robert Forward published work showing how a "vacuum-fluctuation battery" could be constructed. The battery can be recharged by making the electrical forces slightly stronger than the Casimir force to reexpand the plates. (so it is more of an advanced capacitor or rechargable battery than it is a power source)

In 1995 and 1998 Maclay et al. published the first models of a microelectromechanical system (MEMS) with Casimir forces. While not exploiting the Casimir force for useful work, the papers drew attention from the MEMS community due to the revelation that Casimir effect needs to be considered as a vital factor in the future design of MEMS. In particular, Casimir effect might be the critical factor in the stiction failure of MEMS.

In 1999 Pinto, a former scientist at NASA's Jet Propulsion Laboratory at Caltech in Pasadena, published in Physical Review his Gedankenexperiment for a "Casimir engine". The paper showed that continuous positive net exchange of energy from the Casimir effect was possible, even stating in the abstract "In the event of no other alternative explanations, one should conclude that major technological advances in the area of endless, by-product free-energy production could be achieved." In 2001 Capasso et al. showed how the force can be used to control the mechanical motion of a MEMS device, The researchers suspended a polysilicon plate from a torsional rod – a twisting horizontal bar just a few microns in diameter. When they brought a metallized sphere close up to the plate, the attractive Casimir force between the two objects made the plate rotate. They also studied the dynamical behaviour of the MEMS device by making the plate oscillate. The Casimir force reduced the rate of oscillation and led to nonlinear phenomena, such as hysteresis and bistability in the frequency response of the oscillator. According to the team, the system’s behaviour agreed well with theoretical calculations.

Despite this and several similar peer reviewed papers, there is not a consensus as to whether such devices can produce a continuous output of work. Garret Moddel at University of Colorado has highlighted that he believes such devices hinge on the assumption that the Casimir force is a nonconservative force, he argues that there is sufficient evidence (e.g. analysis by Scandurra (2001)) to say that the Casimir effect is a conservative force and therefore even though such an engine can exploit the Casimir force for useful work it cannot produce more output energy then has been input into the system.

In 2008 DARPA solicited research proposals in the area of Casimir Effect Enhancement (CEE). The goal of the program is to develop new methods to control and manipulate attractive and repulsive forces at surfaces based on engineering of the Casimir Force.

A 2008 patent by Haisch and Moddel details a device that is able to extract power from zero-point fluctuations using a gas that circulates through a Casimir cavity. As gas atoms circulate around the system they enter the cavity. Upon entering the electrons spin down to release energy via electromagnetic radiation. This radiation is then extracted by an absorber. On exiting the cavity the ambient vacuum fluctuations (i.e. the zero-point field) impart energy on the electrons to return the orbitals to previous energy levels, as predicted by Senitzky (1960). The gas then goes through a pump and flows through the system again. A published test of this concept by Moddel was performed in 2012 and seemed to give excess energy that could not be attributed to another source. However it has not been conclusively shown to be from zero-point energy and the theory requires further investigation.

(ed note: see original article for links to references)

From the Wikipedia entry for ZERO-POINT ENERGY

8.19 The vacuum energy drive

     The most powerful theories in physics today are quantum theory and the theories of special and general relativity. Unfortunately, those theories are not totally consistent with each other. If we calculate the energy associated with an absence of matter—the "vacuum state"—we do not, as common sense would suggest, get zero. Instead, quantum theory assigns a specific energy value to a vacuum.

     In classical thinking, one could argue that the zero point of energy is arbitrary, so we could simply start measuring energies from the vacuum energy value. However, if we accept general relativity that option is denied to us. Energy, of any form, produces spacetime curvature, and we are therefore not allowed to redefine the origin of the energy scale. Once this is accepted, the energy of the vacuum cannot be talked out of existence. It is real, and when we calculate it we get a large positive value per unit volume.

     How large?

     Richard Feynman addressed the question of the vacuum energy value and computed an estimate for the equivalent mass per unit volume. The estimate came out as two billion tons per cubic centimeter. The energy in two billion tons of matter is more than enough to boil all Earth's oceans.

     Is there any possibility that the vacuum energy could be tapped for useful purposes? Robert Forward has proposed a mechanism, based upon a real physical phenomenon known as the Casimir Effect. I think it would work, but the energy produced is small. The well-publicized mechanisms of others, such as Harold Puthoff, for extracting vacuum energy leave me totally unpersuaded.

     Science fiction that admits it is science fiction is another matter. According to Arthur Clarke, I was the first person to employ the idea of the vacuum energy drive in fictional form, in the story "All the Colors of the Vacuum" (Sheffield, 1981). Clarke employed one in The Songs of Distant Earth (Clarke, 1986). Not surprisingly, there was a certain amount of hand-waving on both Clarke's part and mine as to how the vacuum energy drive was implemented. If the ship can obtain energy from the vacuum, and mass and energy are equivalent, why can't the ship get the reaction mass, too? How does the ship avoid being slowed when it takes on energy, which has an equivalent mass that is presumably at rest? If the vacuum energy is the energy of the ground state, to what new state does the vacuum go, after energy is extracted?

     Good questions. Look on them as an opportunity. There must be good science-fictional answers to go with them.

From BORDERLANDS OF SCIENCE by Charles Sheffield (1999)

      McAndrew laughed, a humorless bark. "I'll tell you why, Jeanie. You flew the Merganser. Tell me how the drive worked."

     "Well, the mass plate at the front balanced the acceleration, so we didn't get any sensation of fifty gee." I shrugged. "I didn't work out the math for myself, but I'm sure I could have if I felt like it."

     I could have, too. I was a bit rusty, but you never lose the basics once you have them planted deep enough in your head.

     "I don't mean the balancing mechanism, that was just common sense." He shook his head. "I mean the drive. Didn't it occur to you that we were accelerating a mass of trillions of tons at fifty gee? If you work out the mass conversion rate you will need, you find that even with an ideal photon drive you'll consume the whole mass in a few days. The Merganser got its drive by accelerating charged particles up to within millimeters a second of light speed. That was the reaction mass. But how did it get the energy to do it?"

     I felt like telling him that when I had been on Merganser there had been other details—such as survival—on my mind. I thought for a few moments, then shook my head.

     "You can't get more energy out of matter than the rest mass energy, I know that. But you're telling me that the drives on Merganser and Hoatzin do it. That Einstein was wrong."

     "No!" McAndrew looked horrified at the thought that he might have been criticizing one of his senior idols. "All I've done is build on what Einstein did. Look, you've done a fair amount of quantum mechanics. You know that when you calculate the energy for the vacuum state of a system you don't get zero. You get a positive value."

     I had a hazy recollection of a formula swimming back across the years. What was it? h/4πw, said a distant voice.

     "But you can set that to zero!" I was proud at remembering so much. "The zero point of energy is arbitrary."

     "In quantum theory it is. But not in general relativity." McAndrew was beating back my mental defenses. As usual when I spoke with him on theoretical subjects, I began to feel I would know less at the end of the conversation than I did at the beginning.

     "In general relativity," he went on, "energy implies space-time curvature. If the zero-point energy is not zero, the vacuum self-energy is real. It can be tapped, if you know what you are doing. That's where Hoatzin draws its energy. The reaction mass it needs is very small. You can get that by scooping up matter as you go along, or if you prefer it you can use a fraction—a very small fraction—of the mass plate."

From ALL THE COLORS OF THE VACUUM by Charles Sheffield (1981)

The first suggestion that vacuum energies might be used for propulsion appears to have been made by Shinichi Seike in 1969. (‘Quantum electric space vehicle’; 8th Symposium on Space Technology and Science, Tokyo.)

Ten years later, H. D. Froning of McDonnell Douglas Astronautics introduced the idea at the British Interplanetary Society’s Interstellar Studies Conference, London (September 1979) and followed it up with two papers: ‘Propulsion Requirements for a Quantum Interstellar Ramjet’ (JBIS, Vol. 33,1980) and ‘Investigation of a Quantum Ramjet for Interstellar Flight’ (AIAA Preprint 81-1534, 1981).

Ignoring the countless inventors of unspecified ‘space drives,’ the first person to use the idea in fiction appears to have been Dr Charles Sheffield, Chief Scientist of Earth Satellite Corporation; he discusses the theoretical basis of the ‘quantum drive’ (or, as he has named it, ‘vacuum energy drive’) in his novel The McAndrew Chronicles (Analog magazine 1981; Tor, 1983).

An admittedly naive calculation by Richard Feynman suggests that every cubic centimetre of vacuum contains enough energy to boil all the oceans of Earth. Another estimate by John Wheeler gives a value a mere seventy-nine orders of magnitude larger. When two of the world’s greatest physicists disagree by a little matter of seventy-nine zeros, the rest of us may be excused a certain scepticism; but it’s at least an interesting thought that the vacuum inside an ordinary light bulb contains enough energy to destroy the galaxy … and perhaps, with a little extra effort, the cosmos.

In what may hopefully be an historic paper (‘Extracting electrical energy from the vacuum by cohesion of charged foliated conductors,’ Physical Review, Vol. 30B, pp. 1700-1702, 15 August 1984) Dr Robert L. Forward of the Hughes Research Labs has shown that at least a minute fraction of this energy can be tapped. If it can be harnessed for propulsion by anyone besides science-fiction writers, the purely engineering problems of interstellar — or even intergalactic — flight would be solved.

From THE SONGS OF DISTANT EARTH by Sir Arthur C. Clarke (1985)

Ladderdown Reactors

Ladderdown transmutation reactors are fringe science invented by Wil McCarthy for his science fiction novel Bloom. It is certainly nothing we will be capable of making anytime soon, but it will take somebody more knowledgeable than me to prove it impossible. Offhand I do not see anything that straight out violates the laws of physics. Ladderdown is unobtainium, not handwavium

Basically ladderdown reactors obtain their energy the same way nuclear fission does: by splitting atomic nuclei and releasing the binding energy. It is just that the ladderdown reactor can work with any element heavier than Iron-56, and the splitting does not release any neutrons or gamma radiation. Nuclear fission only works with fission fuel, and any anti-nuclear activist can tell you horror stories about the dire radiation produced.

Apparently ladderdown reactors remove protons and neutrons from the fuel material one at a time, by quantum tunneling, quietly. Unlike fission, which shoots neutrons like bullets at nuclei, shattering the nucleus into sprays of radiation and exploding fission products.

As with fission the laddered-down nuclei releases the difference in binding energy and moves down the periodic table. The process comes to a screeching halt when the fuel transmutes into Iron-56, since it is at the basin of the binding energy curve. In the novel iron is the most worthless element for this reason, and so is used for cheap building material.

Ladderdown reactors can also take fuel elements that are lighter than Iron-56, and add protons and neutrons one at a time, to make heavier elements (called "ladderup"). This is the ladderdown version of fusion, except it will work with any element lighter than Iron-56 and there is no nasty radiation produced. This is handy because laddering down heavy elements produces lots of protons as a by product, which can be laddered up into Iron-56.


"Now that we are dependent on heavy metals rather than fossil organics and sunlight, economics have simply gone away. You want a lesson in economics from a biophysicist's point of view? It works like ecology—it breeds and selects. Not that we actually carry them in our pockets, but the gram of uranium has become our most basic unit of currency. Thanks to chronic short-staffing, we consider it equivalent to half an hour of human labor, though its energy potential is some twenty-six million times greater. Aside from ourselves, it is the first driver of our economy, the reasons for which are not at all arbitrary."

"For energy reasons," I said.

He winced slightly, shifted position in his chair. "Energy? Well, yes and no. Energy is less important than transmutation potential. In rough terms, a fusion reactor cascading a gram of deuterium/tritium up into a gram of iron—the basin of the binding energy curve—will liberate enough energy to boil about twenty thousand tons of water. A gram of uranium in a ladderdown reactor produces approximately the same. And yet, the uranium is worth ten thousand times more, because in laddering it down, we don't have to sink all the way to iron. We can stop anywhere along the way, and our waste products are isotopes of hydrogen which we can cascade back up, again stopping wherever we like below that magic number, iron fifty-six. A ladderdown economy sees value not only in what a substance is, but also in what it can become, and uranium, alone among the stable elements, can become anything." (all the elements with an atomic number over 82 {uranium} only have isotopes that are known to decompose through radioactive decay.)

Many people are surprised to learn that lead's energy potential is only twenty-five percent less than uranium's, but the thing to remember is that lead has ten fewer transmutation targets—eighty-one versus ninety-one—which translates into a factor of a thousand reduction in its value (Lead has 82 protons and uranium has 92 protons, so lead has 10 fewer transmutation targets). Gold, three rungs lower still, is worth about a five-thousandth as much as uranium (Gold has 79 proton, 13 fewer transmutation targets than uranium, 3 fewer than lead). It has beautiful mechanical and electrical properties, but really, the major cost of paving the streets with it is the labor.

The energy density of antihydrogen is about 250 times what we can achieve with ladderdown, and the production and storage are difficult. Wonderful fuel, the best, but the last time I checked, a gram of it cost over eighty thousand g.u.

Turns out we'll be paying for our food and clothing purchases after all, using, of all things, our shoes. No kidding! Our guide pointed out some bracelets, and though they were fashioned of plain gold he assured us they were very expensive. From the labor that went into them, I assumed, for they were handmade, but no, it turns out the fingernail-sized "dollars" that have been spent on our behalf are also made of gold, and derive their value from their own intrinsic worth as metal. As if we Munies walked around trading actual grams of uranium back and forth. This is what comes of not using ladderdown!

The Gladholders think our "duck shoes" are frightfully amusing anyway, and when they found out what the sole weights were made of, I thought they'd never stop laughing. and when they offered to replace those same blocks with equivalent masses of lead, which of course is five times more valuable back home, I thought we'd never stop laughing.

The biophysicist's voice came back careful, almost embarrassed. "It's never been a secret that nuclear energy presents … certain dangers. We think of ladderdown as a 'clean' technology, which in a radiation sense it certainly is."


"But. The quantum spatial distortion is normally induced and focused within a shielded reactor, where its effects can be controlled to within a few Planck radii. How else to tunnel out only the desired nucleons, yes? But if we invert the distortion function along the B-axis, essentially turning it inside out in three-dimensional space, the same ladderdown tunneling can be induced stochastically in a much larger spherical shell, centered about the inductor. Shielding irrelevant, because it's inside the affected region, you see? Considered too hazardous for use in bloom cauterization, the phenomenon has no industrial applications. Look it up under Things Not to Try."

Most of that went right over my head, but the gist seemed clear enough: he was talking about releasing energy, lots of it, in an uncontrolled manner. He was talking about a bomb.

From BLOOM by Wil McCarthy (1998)

Mass Converters

Mass Converters are fringe science. You see them in novels like Heinlein's Farmer in the Sky, James P. Hogan's Voyage from Yesteryear, and Vonda McIntyre's Star Trek II: The Wrath of Khan. You load the hopper with anything made of matter (rocks, raw sewage, dead bodies, toxic waste, old AOL CD-ROMS, belly-button lint, etc.) and electricity comes out the other end. In the appendix to the current edition of Farmer in the Sky Dr. Jim Woosley is of the opinion that the closest scientific theory that would allow such a thing is Preon theory.

Preon theory was all the rage back in the 1980's, but it seems to have fallen into disfavor nowadays (due to the unfortunate fact that the Standard Model gives better predictions, and absolutely no evidence of preons has ever been observed). Current nuclear physics holds that all subatomic particles are either leptons or composed of groups of quarks. The developers of Preon theory thought that two classes of elementary particles does not sound very elementary at all. So they theorized that both leptons and quarks are themselves composed of smaller particles, pre-quarks or "preons". This would have many advantages.

One of the most complete Preon theory was Dr. Haim Harari's Rishon model (1979). The point of interest for our purposes is that the sub-components of electrons, neutrons, protons, and electron anti-neutrinos contain precisely enough rishon-antirishon pairs to completely annihilate. All matter is composed of electrons, neutrons, and protons. Thus it is theoretically possible in some yet as undiscovered way to cause these rishons and antirishons to mutually annihilate and thus convert matter into energy.

Both James P. Hogan and Vonda McIntyre new a good thing when they saw it, and quickly incorporated it into their novels.

Back about the same time, when I was a young man, I thought I had come up with a theoretical way to make a mass converter. Unsurprisingly it wouldn't work. My idea was to use a portion of antimatter as a catalyst. You load in the matter, and from the antimatter reserve you inject enough antimatter to convert all the matter into energy. Then feed half (or a bit more than half depending upon efficiency) into your patented Antimatter-Makertm and replenish the antimatter reserve. The end result was you fed in matter, the energy of said matter comes out, and the antimatter enables the reaction but comes out unchanged (i.e., the definition of a "catalyst").

Problem #1 was that pesky Law of Baryon Number Conservation, which would force the Antimatter-Maker to produce equal amounts of matter and antimatter. Which would mean that either your antimatter reserve would gradually be consumed or there would be no remaining energy to be output, thus ruining the entire idea. Drat!

Problem #2 is that while electron-positron annihilation produces 100% of the energy in the form of gamma-rays, proton-antiproton annihilation produces 70% as energy and 30% as worthless muons and neutrinos.

Pity, it was such a nice idea too. If you were hard up for input matter, you could divert energy away from the Antimatter-maker and towards the output. Your antimatter reserve would diminish, but if you found more matter later you could run the mass converter and divert more energy into the Antimatter-maker. This would replenish your reserve. And if you somehow totally ran out of antimatter, if another friendly ship came by it could "jump-start" you by connecting its mass converter energy output directly to your Antimatter-maker and run it until you had a good reserve.

Power Storage

Often the power plant generates more power than is currently needed. Spacecraft cannot afford to throw the excess power away, it has to be stored for later use. This is analogous to Terran solar power plants, they don't work at night so you have to store some power by day.

Energy Transport Mechanism

There are a couple of instances where people make the mistake of labeling something a "power source" when actually it is an "energy transport mechanism." The most common example is hydrogen. Let me explain.

In the so-called "hydrogen economy", proponents point out how hydrogen is a "green" fuel, unlike nasty petroleum or gasoline. Burn gasoline and in addition to energy you also produce toxic air pollution. Burn hydrogen and the only additional product is pure water.

The problem is they are calling the hydrogen a fuel, which it isn't.

While there do exist petroleum wells, there ain't no such thing as a hydrogen well. You can't find hydrogen just lying around somewhere, the stuff is far too reactive. Hydrogen has to be generated by some other process, which consumes energy (such as electrolysing water using electricity generated by a coal-fired power plant). Not to mention the energy cost of compressing the hydrogen into liquid, transporting the liquid hydrogen in a power-hungry cryogenically cooled tank, and the power required to burn it and harvest electricity.

This is why hydrogen is not a fuel, it is an energy transport mechanism. It is basically being used to transport the energy from the coal-fired power plant into the hydrogen burning automobile. Or part of the energy, since these things are never 100% efficient.

In essence, the hydrogen is fulling much the same role as the copper power lines leading from a power plant to a residential home. It is transporting the energy from the plant to the home. Or you can look at the hydrogen as sort of a rechargable battery, for example as used in a regenerative fuel cell. But one with rather poor efficiency.

The main example from science fiction is antimatter "fuel." Unless the science fiction universe contains antimatter mines, it is an energy transport mechanism with a truly ugly efficency.


Buck Kendall has invented a sort of super-battery that will store huge amounts of electricity with incredible efficiency. It stores the power in pools of mercury.

"That's it, Tom. I wanted to show you first what we have, and why I wanted all that mercury. Within three weeks, every man, woman and child in the system will be clamoring for mercury metal. That's the perfect accumulator." Quickly he demonstrated the machine, charging it, and then discharging it. It was better than 99.95% efficient on the charge, and was 100% efficient on the discharge.

"Physically, any metal will do. Technically, mercury is best for a number of reasons. It's a liquid. I can, and do it in this, charge a certain quantity, and then move it up to the storage tank. Charge another pool, and move it up. In discharge, I can let a stream flow in continuously if I required a steady, terrific drain of power without interruption. If I wanted it for more normal service, I'd discharge a pool, drain it, refill the receiver, and discharge a second pool. Thus, mercury is the metal to use.

"Do you see why I wanted all that metal?"

"I do, Buck — Lord, I do," gasped Faragaut. "That is the perfect power supply."

"No, confound it, it isn't. It's a secondary source. It isn't primary. We're just as limited in the supply of power as ever — only we have increased our distribution of power."

From THE ULTIMATE WEAPON by John W. Campbell, jr. (1966)


What is needed are so-called "secondary" batteries, commonly known as "rechargable" batteries. If the batteries are not rechargable then they are worthless for power storage. As you probably already figured out, "primary" batteries are the non-rechargable kind; like the ones you use in your flashlight until they go dead, then throw into the garbage.

Current rechargable batteries are heavy, bulky, vulnerable to the space environment, and have a risk of bursting into flame. Just ask anybody who had their laptop computer unexpectedly do an impression of an incindiary grenade.

Nickle-Cadmium and Nickle-Hydrogen rechargables have a specific energy of 24 to 35 Wh/kg (0.086 to 0.13 MJ/kg), an energy density of 0.01 to 0.08 Wh/m3, and an operating temperature range of -5 to 30°C. They have a service life of more than 50,000 recharge cylces, and a mission life of more than 10 years. Their drawbacks are being heavy, bulky, and a limited operationg temperature range.

Lithium-Ion rechargables have a specfic energy of 100 Wh/kg (0.36 MJ/kg), an energy density of 0.25 Wh/m3, and an operating temperature range of -20 to 30°C. They have a service life of about 400 recharge cylces, and a mission life of about 2 years. Their drawbacks are the pathetic service and mission life.


A flywheel is a rotating mechanical device that is used to store rotational energy. In a clever "two-functions for the mass-price of one" bargain a flyweel can also be used a a momentum wheel for attitude control. NASA adores these bargains because every gram counts.

Flywheels have a theoretical maximum specific energy of 2,700 Wh/kg (9.7 MJ/kg). They can quickly deliver their energy, can be fully discharged repetedly without harm, and have the lowest self-discharge rate of any known electrical storage system. NASA is not currently using flywheels, though they did have a prototype for the ISS that had a specific energy of 30 Wh/kg (0.11 MJ/kg).

Regenerative Fuel Cells

A "regenerative" or "reverse" fuel cell is one that saves the water output, and uses a secondary power source (such as a solar power array) to run an electrolysers to split the water back into oxygen and hydrogen. This is only worth while if the mass of the secondary power source is low compared to the mass of the water. But it is attractive since most life support systems are already going to include electrolysers anyway.

In essence the secondary power source is creating fuel-cell fuel as a kind of battery to store power. It is just that a fuel cell is required to extract the power from the "battery."

Currently there exist no regenerative fuel cells that are space-rated. The current goal is for such a cell with a specific energy of up to 1,500 Wh/kg (5.4 MJ/kg), a charge/discharge efficiency up to 70%, and a service life of up to 10,000 hours.

Superconducting magnetic energy storage

Superconducting Magnetic
Energy Storage
Specific energy4–40 kJ/kg
(0.004–0.04 MJ/kg)
(1–11 Wh/kg)
Energy densityless than 40 kJ / L
Specific power10–100,000 kW/kg
Self-discharge rate>0% at 4 K
100% at 140 K
Cycle durabilityUnlimited cycles

Superconducting Magnetic Energy Storage (SMES) systems store energy in the magnetic field created by the flow of direct current in a superconducting coil which has been cryogenically cooled to a temperature below its superconducting critical temperature.

A typical SMES system includes three parts: superconducting coil, power conditioning system and cryogenically cooled refrigerator. Once the superconducting coil is charged, the current will not decay and the magnetic energy can be stored indefinitely.

The stored energy can be released back to the network by discharging the coil. The power conditioning system uses an inverter/rectifier to transform alternating current (AC) power to direct current or convert DC back to AC power. The inverter/rectifier accounts for about 2–3% energy loss in each direction. SMES loses the least amount of electricity in the energy storage process compared to other methods of storing energy. SMES systems are highly efficient; the round-trip efficiency is greater than 95%.

Due to the energy requirements of refrigeration and the high cost of superconducting wire, SMES is currently used for short duration energy storage. Therefore, SMES is most commonly devoted to improving power quality.

Low-temperature versus high-temperature superconductors

Under steady state conditions and in the superconducting state, the coil resistance is negligible. However, the refrigerator necessary to keep the superconductor cool requires electric power and this refrigeration energy must be considered when evaluating the efficiency of SMES as an energy storage device.

Although the high-temperature superconductor (HTSC) has higher critical temperature, flux lattice melting takes place in moderate magnetic fields around a temperature lower than this critical temperature. The heat loads that must be removed by the cooling system include conduction through the support system, radiation from warmer to colder surfaces, AC losses in the conductor (during charge and discharge), and losses from the cold–to-warm power leads that connect the cold coil to the power conditioning system. Conduction and radiation losses are minimized by proper design of thermal surfaces. Lead losses can be minimized by good design of the leads. AC losses depend on the design of the conductor, the duty cycle of the device and the power rating.

The refrigeration requirements for HTSC and low-temperature superconductor (LTSC) toroidal coils for the baseline temperatures of 77 K, 20 K, and 4.2 K, increases in that order. The refrigeration requirements here is defined as electrical power to operate the refrigeration system. As the stored energy increases by a factor of 100, refrigeration cost only goes up by a factor of 20. Also, the savings in refrigeration for an HTSC system is larger (by 60% to 70%) than for an LTSC systems.

From the Wikipedia entry for

There are two significant limits.

First is the force trying to make the superconductor explode.

You can consider that the energy of a persistent supercurrent circulating through a superconductor is stored in the magnetic field it produces. The best design is thus to wrap your superconducting wire into a solenoid (or inductor, or electromagnet). To avoid annoying effects from the extremely strong field leaking out of the end, wrap the ends of the solenoid around so they join, giving a torroidal (or doughnut shaped) configuration. Now the field will act to maintain the current that produces it, and can induce strong "voltages" (technically an electromotive force, or EMF, but for practical purposes you can treat it as a voltage from a battery) to drive the current through any load you apply to it.

But now you have a problem. The current produces the field, and you need the field to maintain the current. But the field also exerts a force on the current, pushing the current-carrying loops apart and trying to expand them. For high currents and strong field (what you get when you are storing lots of energy), these forces can be high enough to rip matter apart and make your superconductive "battery" explode.

The way to avoid this is to support the superconductive wire with a very strong backing material that holds it in place. The upper limit on the energy storage per unit weight comes down, ultimately, to the strength of the chemical bonds that hold your backing material together. The best you can do here is use some strongly-bound light element. The carbon-carbon chemical bond is going to be ideal. So a carbon super-material like carbon nanotubes or graphene will give you the best energy per weight. The theoretical upper limit is around 40 to 50 MJ/kg (11,000 to 14,000 Wh/kg). Of course, a power storage unit energized up to this limit will be on the verge of failure, and failure means exploding with ten times its weight of TNT (4.2 MJ/kg). So throw in some engineering safety factors of 2 or 3 (25 to 17 MJ/kg or 7,000 to 5,000 Wh/kg).

The other limit is that high enough magnetic fields will shut down a superconductor. This is called the critical field, here the substance goes back to being a normal conductor instead of a superconductor (with the subsequent loss of all of your energy to resistive heating and probably exploding, again). This puts an upper limit on your energy per volume rather than energy per mass. I'm not aware of any theoretical upper limit on what the critical field could be, so you can probably adjust this to whatever you need. Note that you need a safety factor here, too, since the critical field decreases as temperature increases. You don't want your power supply to turn into a bomb just because the air conditioning starts acting up.

From a Google Plus thread entry Luke Campbell (2017)

Kerr-Newman black hole

The popular conception of a black hole is that it sucks everything in, and nothing gets out. However, it is theoretically possible to extract energy from a black hole, for certain values of "from."

And by the way, there appears to be no truth to the rumor that Russian astrophysicists use a different term, since "black hole" in the Russian language has a scatological meaning. It's an urban legend, I don't care what you read in Dragon's Egg.

For an incredibly dense object with an escape velocity higher than the speed of light which warps the very fabric of space around them, black holes are simple objects. Due to their very nature they only have three characteristics: mass, spin (angular momentum), and electric charge. All the other characteristics got crushed away (well, technically they also have magnetic moment, but that is uniquely determined by the other three). All black holes have mass, but some have zero spin and others have zero charge.

There are four types of black holes. If it only has mass, it is a Schwarzschild black hole. If it has mass and charge but no spin, it is a Reissner-Nordström black hole. If it has mass and spin but no charge it is a Kerr black hole. And if it has mass, charge and spin it is a Kerr-Newman black hole. Since practically all natural astronomical objects have spin but no charge, all naturally occurring black holes are Kerr black holes, the others do not exist naturally. In theory one can turn a Kerr black hole into a Kerr-Newman black hole by shooting charged particles into it for a few months, say from an ion drive or a particle accelerator.

From the standpoint of extracting energy, the Kerr-Newman black hole is the best kind, since it has both spin and charge. In his The MacAndrews Chronicles, Charles Sheffield calls them "Kernels" actually "Ker-N-el", which is shorthand for Kerr-Newman black hole.

The spin acts as a super-duper flywheel. You can add or subtract spin energy to the Kerr-Newman black hole by using the Penrose process. Just don't extract all the spin, or the blasted thing turns into Reissner-Nordström black hole and becomes worthless. The attractive feature is that this process is far more efficient than nuclear fission or thermonuclear fusion. And the stored energy doesn't leak away either.

The electric charge is so you can hold the thing in place using electromagnetic fields. Otherwise there is no way to prevent it from wandering thorough your ship and gobbling it up.

The assumption is that Kerr-Newman black holes of manageable size can be found naturally in space, already spun up and full of energy. If not, they can serve as a fantastically efficient energy transport mechanism.

Primordial black holes

Alert readers will have noticed the term "manageable size" above. It is impractical to use a black hole with a mass comparable to the Sun. Your ship would need an engine capable of moving something as massive as the Sun, and the gravitational attraction of the black hole would wreck the solar system. So you just use a smaller mass black hole, right? Naturally occurring small black holes are called "Primordial black holes."

Well, there is a problem with that. In 1975 legendary physicist Stephen Hawking discovered the shocking truth that black holes are not black (well, actually the initial suggestion was from Dr. Jacob Bekenstein). They emit Hawking radiation, for complicated reasons that are so complicated I'm not going to even try and explain them to you (go ask Google). The bottom line is that the smaller the mass of the black hole, the more deadly radiation it emits. The radiation will be the same as a "black body" with a temperature of:

6 × 10-8 / M kelvins

where "M" is the mass of the black hole where the mass of the Sun equals one. The Sun has a mass of about 1.9891 × 1030 kilograms.

Jim Wisniewski created an online Hawking Radiation Calculator to do the math for you.

In The McAndrew Chronicles Charles Sheffield hand-waved an imaginary force field that somehow contained all the deadly radiation. One also wonders if there is some way to utilze the radiation to generate power.

In the table:

  • R is the black hole's radius in attometers (units of one-quintillionth or 10-18 of a meter). A proton has a diameter of 1000 attometers.
  • M is the mass in millions of metric tons. One million metric tons is about the mass of three Empire State buildings.
  • kT is the Hawking temperature in GeV (units of one-billion Electron Volts).
  • P is the estimated total radiation output power in petawatts (units of one-quadrillion watts). 1—100 petawatts is the estimated total power output of a Kardashev type 1 civilization.
  • P/c2 is the estimated mass-leakage rate in grams per second.
  • L is the estimated life expectancy of the black hole in years. 0.04 years is about 15 days. 0.12 years is about 44 days.

Table is from Are Black Hole Starships Possible?, thanks to magic9mushroom for this link.

"I think Earth's worst problems are caused by the power shortage," he said. "That affects everything else. Why doesn't Earth use the kernels for power, the way that the USF does?"

"Too afraid of an accident," replied McAndrew. His irritation evaporated immediately at the mention of his specialty. "If the shields ever failed, you would have a Kerr-Newman black hole sitting there, pumping out a thousand megawatts—mostly as high-energy radiation and fast particles. Worse than that, it would pull in free charge and become electrically neutral. As soon as that happened, there'd be no way to hold it electromagnetically. It would sink down and orbit inside the Earth. We couldn't afford to have that happen."

"But couldn't we use smaller kernels on Earth?" asked Yifter. "They would be less dangerous."

McAndrew shook his head. "It doesn't work that way. The smaller the black hole, the higher the effective temperature and the faster it radiates. You'd be better off with a much more massive black hole. But then you've got the problem of supporting it against Earth's gravity. Even with the best electromagnetic control, anything that massive would sink down into the Earth."

"I suppose it wouldn't help to use a nonrotating, uncharged hole, either," said Yifter. "That might be easier to work with."

"A Schwarzschild hole?" McAndrew looked at him in disgust. "Now, Mr. Yifter, you know better than that." He grew eloquent. "A Schwarzschild hole gives you no control at all. You can't get a hold of it electromagnetically. It just sits there, spewing out energy all over the spectrum, and there's nothing you can do to change it—unless you want to charge it and spin it up, and make it into a kernel. With the kernels, now, you have control."

I tried to interrupt, but McAndrew was just getting warmed up. "A Schwarzschild hole is like a naked flame," he went on. "A caveman's device. A kernel is refined, it's controllable. You can spin it up and store energy, or you can use the ergosphere to pull energy out and spin it down. You can use the charge on it to move it about as you want. It's a real working instrument—not a bit of crudity from the Dark Ages."

from THE McANDREW CHRONICLES by Charles Sheffield (1983)

In this model of the interaction of a miniature black hole with the vacuum, the black hole emits radiation and particles, as though it had a temperature. The temperature would be inversely proportional to the mass of the black hole. A Sun-sized black hole is very cold, with a temperature of about a millionth of a degree above absolute zero. When the mass of the black hole is about a hundred billion tons (the mass of a large asteroid), the temperature is about a billion degrees.

(ed note: one hundred billion tons is 100,000 million tons or 5 × 10-17 solar masses. 6 × 10-8 / 5 × 10-17 = 1,200,000,000 Kelvin)

According to Donald Page, who carried out lengthy calculations on the subject, such a hole should emit radiation that consists of approximately 81% neutrinos, 17% photons, and 2% gravitons. When the mass becomes significantly less than a hundred billion tons, the temperature increases until the black hole is hot enough to emit electrons and positrons as well as radiation. When the mass becomes less than a billion tons (a one kilometer diameter asteroid), the temperature now approaches a trillion degrees and heavier particle pairs, like protons and neutrons are emitted. The size of a black hole with a mass of a billion tons is a little smaller than the nucleus of an atom. The black hole is now emitting 6000 megawatts of energy, the output of a large power plant. It is losing mass at such a prodigious rate that its lifetime is very short and it essentially "explodes" in a final burst of radiation and particles.

(ed note: one billon tons is 1000 million tons. An atomic nucleus is about 1750 to 15,000 attometers in diameter.)

If it turns out that small black holes really do exist, then I propose that we go out to the asteroid belt and mine the asteroids for the black holes that may be trapped in them. If a small black hole was in orbit around the Sun in the asteroid belt region, and it had the mass of an asteroid, it would be about the diameter of an atom. Despite its small size, the gravity field of the miniature black hole would be just as strong as the gravity field of an asteroid and if the miniature black hole came near another asteroid, the two would attract each other. Instead of colliding and fragmenting as asteroids do, however, the miniature black hole would just penetrate the surface of the regular asteroid and pass through to the other side. In the process of passing through, the miniature black hole would absorb a number of rock atoms, increasing its weight and slowing down slightly. An even more drastic slowing mechanism would be the tides from the miniature black hole. They would cause stresses in the rock around the line of penetration and fragment the rock out to a few micrometers away from its path through the asteroid. This would cause further slowing.

After bouncing back and forth through the normal matter asteroid many times, the miniature black hole would finally come to rest at the center of the asteroid. Now that it is not moving so rapidly past them, the miniature black hole could take time to absorb one atom after another into its atom-sized body until it had dug itself a tiny cavity at the center of the asteroid. With no more food available, it would stop eating, and sit there and glow warmly for a few million years. After years of glowing its substance away, it would get smaller. As it got smaller it would get hotter since the temperature rises as the mass decreases. Finally, the miniature black hole would get hot enough to melt the rock around it. Drops of melted rock would be pulled into the miniature black hole, adding to its mass. As the mass of the black hole increased, the temperature would decrease. The black hole would stop radiating, the melted rock inside the cavity would solidify, and the process would repeat itself many centuries later. Thus, although a miniature black hole left to itself has a lifetime that is less than the time since the Big Bang, there could be miniature black holes with the mass of an asteroid, being kept alive in the asteroid belt by a symbiotic interaction with an asteroid made of normal matter.

To find those asteroids that contain miniature black holes, you want to look for asteroids that have anomalously high temperatures, lots of recent fracture zones, and anomalously high density. Those with a suspiciously high average density have something very dense inside. To obtain a measure of the density, you need to measure the volume and the mass. It is easy enough to get an estimate of the volume of the host asteroid with three pictures taken from three different directions. It is difficult to measure the mass of an object in free fall. One way is to go up to it with a calibrated rocket engine and push it. Another is to land on it with a sensitive gravity meter. There is, however, a way to measure the mass of an object at a distance without going through the hazard of a rendezvous. To do this, you need to use a mass detector or gravity gradiometer.

Once you have found a suspiciously warm asteroid that seems awfully massive for its size, then to extract the miniature black hole, you give the surface of the asteroid a strong shove and push the asteroid out of the way. The asteroid will shift to a different orbit, and where the center of the asteroid used to be, you will find the miniature black hole. The black hole will be too small to see, but if you put an acoustic detector on the asteroid you will hear the asteroid complaining as the black hole comes to the surface. Once the black hole has left the surface you can monitor its position and determine its mass with a mass detector.

The next step in corralling the invisible black maverick is to put some electric charge on it. This means bombarding the position of the miniature black hole with a focused beam of ionized particles until the black hole has captured enough of them to have a significant charge to mass ratio. The upper limit will depend upon the energy of the ions. After the first ion is absorbed, the black hole will have a charge and will have a tendency to repel the next ion. Another upper limit to the amount of charge you can place on a black hole is the rate at which the charged black hole pulls opposite charges out of the surrounding space. You can keep these losses low, however, by surrounding the black hole with a metal shield.

Once a black hole is charged, you can apply forces to it with electric fields. If the charged black hole happens to be rotating, you are in luck, for then it will also have a magnetic field and you can also use magnetic fields to apply forces and torques. The coupling of the electric charge to the black hole is very strong—the black hole will not let go. You can now use strong electric or magnetic fields to pull on the black hole and take it anywhere you want to go.

from INDISTINGUISHABLE FROM MAGIC by Robert L. Forward (1995)

(ed note: for you Ugly Americans who have never heard of Perry Rhodan, this is a science fictional device)

Schwarzschild reactors have power output ten thousands time higher than a fusion reactor.

The reactor create a artificial pulsating micro black hole in size of one hundred nanometers. It shifts between being a black hole with event horizon and space time warp with no event horizon.

The black hole is fed with particle beam of ultra-catalyzed deuterium. Approximately 50% of deuterium is transformed into gamma-rays, the rays are collected by "super solar cells" and transformed into usable energy with an efficiency of 80%.

The other 50% of the deuterium is transformed into antimatter, swallowed by black hole (in space time warp mode) where it vanishes into the depths of hyperspace.

Michel Van (2015)

(ed note: for you Ugly Americans who have never heard of Perry Rhodan, this is a science fictional device)

Humans found the Schwarzschild reactors performance to be disappointing. Only 50% deuterium into gamma-rays could be improved upon. Human scientists developed the NUGAS-Schwarzschild Reactors.

The principle remain almost the same.

However instead of the antimatter being discharged into hyperspace, it is directed into the path of a particle beam for mutual annihilation. Thus 100% of the deuterium is converted into gamma rays.

Due the higher pulse rate and antimatter annihilation, ultra catalyzed deuterium was unsuitable as fuel. Instead ionized hydrogen nucleons (protons) were subsituted. They are conpressed to a density of 3.5×107 kilograms per cubic meter to form the Nucleon Gas (NUGAS) fuel ball. The NUGAS fuel ball has a mass of 200,000 metric tons. It is surrounded by containment generators forming a reactor with a diameter of 12 meters.

NUGAS is also used as fuel for starship Puls proton beam engines, the successor to the older impuls engines.

Of course NUGAS is dangerous, but it gave the 1970s Perry Rhodan authors interesting plot complications (such as a NUG-ball in danger of losing its containment field). The technological levels in the Perry Rhodan universe eventually became too unbelievable, so in 2003 the authors "reset" it to tone everything down. Now NUGAS only compress to a density of 8.75×106 kg/m3, and have a mass of only 50,000 metric tons.

The idea of the Schwarzschild Reactor and the NUG version came from German science fiction author Kurt Mahn. He was a real life physcist who worked for Pratt & Whitney, Martin Marietta, and Harris Electronics. He wrote for Perry Rhodan from 1962 to 1969 and later from 1972 to 1993.

Michel Van (2015)

Heat Radiators

RocketCat sez

You might think that the problem with surviving in the "zero degree cold of space" is keeping from freezing to death. Nope, the problem is Heat.

Human bodies are little furnaces, which you can discover if you wrap your limbs and torso in plastic wrap and see how little time it takes to pass out from heat prostration (note to jackasses THIS IS AN ILLUSTRATION, DO NOT ACTUALLY TRY TO DO THIS!). In space, it's not like you can open the window for a cooling breeze, either. Your cosy little habitat module will turn into an oven.

If you stayed awake during Physics 101 class you'll know that the blasted laws of thermodynamics say there are only three ways of getting rid of waste heat. But only one of them will work in space: radiation.

So you'll need heat radiators or the crew is going to die horribly while sweating bullets.

And I am quite sure that you are going to make things infinitely worse by insisting on your precious nuclear power reactors and megawatt laser cannons. Human bodies only make enough waste heat to kill everybody, reactors and lasers can make the entire freaking ship glow white-hot and vaporize.

Ever see those titanic curved towers around nuclear power plants? Yep, cooling towers. You'll need something a bit more high-tech if you do not want your spacecraft's aesthetics spoiled by a 40 meter cooling tower or two.

Laser cannon are much worse. Rick Robinson described them as observatory telescopes with a jet engine at the eyepiece. Ken Burnside said they were blast furnaces that produced coherent light as a byproduct. Whatever you call them they are hot enough to make your ship go from solid directly to Solar-surface hot ionized gas without passing through the molten metal stage first.

But of course heat radiators are one of the major things conspicuous by their absence in science fiction TV shows and movies. Concept artists don't want their ultra-futuristic spacecraft decked with 17th century billowed sails. They even over-ruled Arthur C. Clarke for cryin' out loud! The only exception that comes to mind is the ISV Venture Star from the movie Avatar.

Power plants and some propulsion systems are going to require heat radiators to avoid system meltdown. There are only three ways of getting rid of heat: convection, conduction, and radiation; and the first two do not work at all in the vacuum of space. So the ship designer is stuck with heat radiators, or what NASA calls Active Thermal Control Systems

Functionally they are not too different from the radiator on your automobile. Pipes full of radiator fluid are coiled around the cylinder heads and engine block, sucking up the heat so the engine doesn't turn into molten lava. The hot radiator fluid is moved by the coolant pump, carrying the heat into the engine coolant radiator (that flat box on the automobile's nose with all the scalloped holes). In the radiator, the heat is removed from the radiator fluid by conduction with the wind. The cool radiator fluid travels into the engine and the cycle begins anew.

Actually, in spaceships the heat radiators get rid of heat by … well … radiating, instead of conduction. Different design because there is no wind in space. But you get the idea.

See Thermophotovoltaic Energy Conversion in Space Nuclear Reactor Power Systems and HIGH TRADER for details.


If you want to calculate this for yourself use the Stefan-Boltzmann law:

P = A * ε * σ * T4

A = P / (ε * σ * T4)


  • P = the power of waste heat the radiator can get rid of (watts)
  • σ = 5.670373×10-8 = Stefan-Boltzmann constant (W m-2K-4)
  • ε = emissivity of radiator (theoretical maximum is 1.0 for a perfect black body, real world radiator will be less. Should be at least 0.8 or above to be worth-while)
  • A = area of radiator (m2)
  • T = temperature of radiator, this assumes temperature of space is zero degrees (degrees K)
  • x4 = raise x to the fourth power, i.e, x * x * x * x

My source (Matthew DeBell) says that if P = 150 gigawatts, ε = 0.94, and T = 3000 K, A would be 34,941 m2. Actually it could be half that if you have a two-sided radiator, which would make the radiator 17,470 m2 (a square 132 meters on a side). Which is still freaking huge.

For estimating the mass of the radiator array, go here.

Ken Burnside says that if one examine the equation carefully one will notice that the radiator effectiveness goes up at the fourth power of the heat of the radiator. The higher the temperature, the lower the surface area can be, which lowers the required mass of radiator fins. This is why most radiator designs use liquid sodium or lithium (or things more exotic, still). 1600K radiators mean that you need a lot less mass than 273 K radiators.

Ken Burnside also noted that radiators are large, flimsy, and impossible to armor (except perhaps for the droplet radiator). A liability on a warship. However, Zane Mankowski (author of Children of a Dead Earth) makes a good case that heat radiators can indeed be armored. Mr. Mankowski says the thickness of the radiator material can be increased to provide armor-like protection for the working fluid tubes, with the price of reducing radiator efficiency.

Mr. Burnside has an entire essay about the problem of heat on combat spacecraft, entitled The Hot Equations: Thermodynamics and Military SF. Since thermodynamics is one of the most important (and most neglected in science fiction) factors in combat, the essay will repay careful study.

In the military the old bromide is that amateurs talk about battle tactics while professionals talk about logistics. In the real of spacecraft design, @AsteroidEnergy said "Amateurs discuss rockets, professionals discuss heat management."

But do realize that if the spacecraft does indeed have a nuclear propulsion system or something else dangerously radioactive, the radiators must be tapered to keep inside the radiation shadow shield. Or bad things happen.


I had initially thought that the heat from the life-system could be simply dumped by the same radiator system dealing with the multi-gigawatt waste heat from the propulsion system or power system. Richard Bell pointed out that I had not thought the problem through. Due to the difference in the temperatures of the waste heat from life-system and propulsion, unreasonably large amounts of energy will be required to get the low-level life-system heat into a radiator designed to handle high-level propulsion heat. The bottom line is that there will be two separate radiator systems.

Not only are you going to require two separate radiator systems, the one for the modest cooling required by the life-system is liable to have larger radiator surfaces than the one cooling the multi-gigawatt propulsion system. Radiator effectiveness goes up as the fourth power of the heat of the radiator, remember?


Propulsion systems like nuclear thermal rockets do not need heat radiators because the waste heat is carried away by the exhaust plume. In effect, the exhaust is their radiator (the technical term is "Open-Cycle Cooling"). Note this only works if the propulsion system has a high propellant mass flow (called "mdot"). Note that the lower the thrust the lower the mdot. Once the thrust gets too low there is not enough propellant in the exhaust plume for you to use the open-cycle cooling trick.

Electrical powered drives like ion drives will require radiators on their power plants. Fusion drives may or may not require radiators, depending upon whether you can make them open-cycle or not.


It should be pointed out that in a vacuum environment, convection is no longer available and the only mechanism of rejecting heat is radiation. Radiation follows the Stefan-Boltzmann Law

E = σT4

E = the energy rejected
σ = the Stefan-Boltzmann constant, = 5.67 W m-2 K-4
T = the temperature at which the heat is radiated

That is, the total amount of heat radiated is proportional to the surface area of the radiator. And the lower the radiation temperature, the larger the radiator area (and thus the radiator mass, for a given design) must be.

The radiator can only reject heat when the temperature is higher than that of the environment. In space, the optimum radiation efficiency is gained by aiming the radiator at free space. Radiating toward an illuminated surface is less effective, and the radiator must be shielded from direct sunlight.

The rejection of heat at low temperatures, such as would be the case in environmental control and in the thermal management of a materials processing unit, is particularly difficult.

Space-Based Power Generating Systems

Solar photovoltaic systems have a generating capability of up to several hundred kilowatts. The power output range of solar thermal systems is expected to be one hundred to perhaps several hundred kilowatts. While in principle these power systems can be expanded into the megawatt region, the prohibitive demands for collection area and lift capacity would appear to rule out such expansion. Megawatt and multimegawatt nuclear power reactors adapted for the space environment appear to offer a logical alternative.

Solar photovoltaics themselves will not burden the power generating system with a direct heat rejection requirement, since the low energy density of the system requires such a great collection area that it allows rejection of waste radiant energy. However, if these systems are to be employed in low Earth orbit or on a nonterrestrial surface, then a large amount of energy storage equipment will be required to ensure a continuous supply of power (as the devices do not collect energy at night). And the round-trip inefficiencies of even the best energy storage system today will require that a large fraction—perhaps 25 percent—of the electrical power generated must be dissipated as waste heat and at low temperatures.

Solar thermal systems, which include a solar concentrator and a dynamic energy conversion system, are presumed to operate at relatively high temperatures (between 1000 and 2000 K). The efficiencies of the energy conversion system will lie in the range of 15 to perhaps 30 percent. Therefore we must consider rejecting between 70 and 85 percent of the energy collected. In general, the lower the thermal efficiency, the higher the rejection temperature and the smaller the radiating area required. As with solar photovoltaic systems, the inefficiencies of the energy storage system will have to be faced by the heat rejection system, unless high temperature thermal storage is elected.

The current concepts for nuclear power generating systems involve reactors working with relatively low-efficiency energy conversion systems which reject virtually all of the usable heat of the reactor but at a relatively high temperature. Despite the burdens that this low efficiency places on nuclear fuel use, the energy density of nuclear systems is so high that the fuel use factor is not expected to be significant.

In all of these systems the output power used by the production system in environmental control and manufacturing (except for a small fraction which might be stored as endothermic heat in the manufactured product) will have to be rejected at temperatures approaching 300 K.

As an example of the severity of this problem, let us examine the case of a simple nuclear power plant whose energy conversion efficiency from thermal to electric is approximately 10 percent. The plant is to generate 100 kW of useful electricity. The reactor operates at approximately 800 K, and a radiator with emissivity equal to 0.85 would weigh about 10 kg/m2. The thermal power to be dissipated from the reactor would be about 1 MW. From the Stefan Boltzmann Law, the area of the radiator would be about 50 m2 and the mass approximately 500 kg. This seems quite reasonable.

However, we must assume that the electricity generated by the power plant, which goes into life support systems and small-scale manufacturing, would eventually have to be dissipated also, but at a much lower temperature (around 300 K). Assuming an even better, aluminum radiator of about 5 kg/m2, with again an emissivity of 0.85, in this case we find that the area of the low temperature heat rejection component is 256 m2, with a mass approaching 1300 kg.

Using the Stefan-Boltzmann Law,

E1 = 5.67×10-8 W m-2 K-4 (800 K)4
E1 = 5.67×10-8 W m-2 K-4 × 4096×108 K4
E1 = 5.67 W m-2 × 4.10×103
E1 = 23.3 kW m-2

900 kW / 23.3 kW m-2 = 38.6 m2
and 38.6 m2 / 0.85 = 45.4 m2

E2 = 5.67×10-8 W m-2 K-4 (300 K)4
E2 = 5.67×10-8 W m-2 K-4 × 81×108 K4
E2 = 5.67 W m-2 × 81
E2 = 459 W m-2

100 kW / 459 W m-2 = 0.2179×103 m2 = 218 m2
and 218 m2 / 0.85 = 256 m2

Therefore, we can see that the dominant heat rejection problem is not that of the primary power plant but that of the energy that is used in life support and manufacturing, which must be rejected at low temperatures. Using the waste heat from the nuclear power plant for processing may be effective. But, ironically, doing so will in turn require more radiator surface to radiate the lower temperature waste heat.

Heat Rejection Systems

In this section I will deal with systems designed to meet the heat rejection requirements of power generation and utilization. These heat rejection systems may be broadly classified as passive or active, armored or unarmored. Each is expected to play a role in future space systems.

Heat pipes: The first of these, called the “heat pipe,” is conventionally considered the base system against which all others are judged. It has the significant advantage of being completely passive, with no moving parts, which makes it exceptionally suitable for use in the space environment.

For the convenience of the reader, I will briefly describe the operational mechanism of the basic heat pipe. (See figure 36.) The heat pipe is a thin, hollow tube filled with a fluid specific to the temperature range at which it is to operate. At the hot end, the fluid is in the vapor phase and attempts to fill the tube, passing through the tube toward the cold end, where it gradually condenses into the liquid phase. The walls of the tube, or appropriate channels grooved into the tube, are filled with a wick-like material which returns the fluid by surface tension to the hot end, where it is revaporized and recirculated.

Essentially the system is a small vapor cycle which uses the temperature difference between the hot and cold ends of the tube as a pump to transport heat, taking full advantage of the heat of vaporization of the particular fluid.

The fluid must be carefully selected to match the temperature range of operation. For example, at very high temperatures a metallic substance with a relatively high vaporization temperature, such as sodium or potassium, may be used. However, this choice puts a constraint on the low temperature end since, if the fluid freezes into a solid at the low temperature end, operation would cease until the relatively inefficient conduction of heat along the walls could melt it. At low temperatures a fluid with a low vaporization temperature, such as ammonia, might well be used, with similar constraints. The temperature may not be so high as to dissociate the ammonia at the hot end or so low as to freeze the ammonia at the cold end.

With proper design, heat pipes are an appropriate and convenient tool for thermal management in space systems. For example, at modest temperatures, the heat pipe could be made of aluminum, because of its relatively low density and high strength. Fins could be added to the heat pipe to increase its heat dissipation area. The aluminum, in order to be useful, must be thin enough to reduce the mass carried into space yet thick enough to offer reasonable resistance to meteoroid strikes.

A very carefully designed solid surface radiator made out of aluminum has the following capabilities in principle: The mass is approximately 5 kg/m2 with an emissivity of 0.85; the usable temperature range is limited by the softening point of aluminum (about 700 K). At higher temperatures, where refractory metals are needed, it would be necessary to multiply the mass of the radiator per square meter by at least a factor of 3. Nevertheless, from 700 K up to perhaps 900 K, the heat pipe radiator is still a very efficient method of rejecting heat.

A further advantage is that each heat pipe unit is a self-contained machine. Thus, the puncture of one unit does not constitute a single-point failure that would affect the performance of the whole system. Failures tend to be slow and graceful, provided sufficient redundancy.

Pump loop system: The pump loop system has many of the same advantages and is bounded by many of the same limitations associated with the heat pipe radiator. Here heat is collected through a system of fluid loops and pumped into a radiator system similar to conventional radiators used on Earth. It should be pointed out that in the Earth environment the radiator actually radiates very little heat; it is designed to convect its heat. The best known examples of the pump loop system currently used in space are the heat rejection radiators used in the Shuttle. These are the inner structure of the clamshell doors which are deployed when the doors are opened (fig. 37).

Pump loop systems have a unique advantage in that the thermal control system can easily be integrated into a spacecraft or space factory. The heat is picked up by conventional heat exchangers within the spacecraft, the carrier fluid is pumped through a complex system of pipes (extended by fins when deemed effective), and finally the carrier is returned in liquid phase through the spacecraft. In the case of the Shuttle, where the missions are short, additional thermal control is obtained by deliberately dumping fluid.

Since the system is designed to operate at low temperatures, a low density fluid, such as ammonia, may on occasion, depending on heat loading, undergo a phase change. Boiling heat transfer in a low gravity environment is a complex phenomenon, which is not well understood at the present time. Because the system is subjected to meteoroid impact, the basic primary pump loops must be strongly protected.

Despite these drawbacks, pump loop systems will probably be used in conjunction with heat pipe systems as thermal control engineers create a viable space environment. These armored (closed) systems are rather highly developed and amenable to engineering analysis. They have already found application on Earth and in space. A strong technology base has been built up, and there exists a rich literature for the scientist-engineer to draw on in deriving new concepts.

Advanced Radiator Concepts

The very nature of the problems just discussed has led to increased efforts on the part of the thermal management community to examine innovative approaches which offer the potential of increased performance and, in many cases, relative invulnerability to meteoroid strikes. Although I cannot discuss all of these new approaches, I will briefly describe some of the approaches under study as examples of the direction of current thinking.

Improved conventional approaches: The continuing search for ways to improve the performance of heat pipes has already shown that significant improvements in the heat pumping capacity of the heat pipe can be made by clever modifications to the return wick loop. Looking further downline at the problem of deployability, people are exploring flexible heat pipes and using innovative thinking. For example, a recent design has the heat pipes collapsing into a sheet as they are rolled up, the same way a toothpaste tube does. Thus, the whole ensemble may be rolled up into a relatively tight bundle for storing and deploying. However, because the thin-walled pipes are relatively fragile and easily punctured by meteoroids, more redundancy must be provided. The same principles, of course, can be applied to a pump loop system and may be of particular importance when storage limits must be considered. These are only examples of the various approaches taken, and we may confidently expect a steady improvement in the capability of conventional thermal management systems.

The liquid droplet radiator: The basic concept of the liquid droplet radiator is to replace a solid surface radiator by a controlled stream of droplets. The droplets are sprayed across a region in which they radiate their heat; then they are recycled to the hotter part of the system. (See figure 38.)

It was demonstrated some time ago that liquid droplets with very small diameters (about 100 micrometers) are easily manufactured and offer a power-to-mass advantage over solid surface radiators of between 10 and 100. In effect, large, very thin radiator sheets can be produced by the proper dispersion of the droplets. This system offers the potential of being developed into an ultralightweight radiator that, since the liquid can be stored in bulk, is also very compact.

The potential advantages of the liquid droplet radiator can be seen if we consider again the problem that was discussed at the end of the section on heat pipe radiators. We found that a very good aluminum radiator would require 256 m2 and have a mass of nearly 1300 kg to radiate the low temperature waste heat from lunar processing. Using the properties of a liquid droplet radiator and a low density, low vapor pressure fluid such as Dow-Corning 705, a common vacuum oil, we find that, for the same area (which implies the same emissivity), the mass of the radiating fluid is only 24 kg.

Even allowing a factor of 4 for the ancillary equipment required to operate this system, the mass of the radiator is still less than 100 kg.

To achieve efficiency, the designer is required to frame the radiator in a lightweight deployable structure and to provide a means of aiming the droplets precisely so that they can be captured and returned to the system. However, present indications are that the droplet accuracies required (milliradians) are easily met by available technology. Recently, successful droplet capture in simulated 0 g conditions has been adequately demonstrated. An advantage of a liquid droplet radiator is that even a relatively large sheet of such droplets is essentially invulnerable to micrometeoroids, since a striking micrometeoroid can remove at most only a few drops.

The reader may be concerned that the very large surface area of the liquid will lead to immediate evaporation. However, liquids have recently been found that in the range of 300 to 900 K have a vapor pressure so low that the evaporation loss during the normal lifetime of a space system (possibly as long as 30 years) will be only a small fraction of the total mass of the radiator.

Thus, the liquid droplet radiator appears promising, particularly as a low temperature system where a large radiator is required.

Liquid droplet radiators for applications other than 0 g have been suggested. For example, in the lunar environment fluids with low vapor pressures can be used effectively as large area heat dissipation systems for relatively large-scale power plants. We may well imagine that such a system will take on the appearance of a decorative fountain, in which the fluid is sprayed upward and outward to cover as large an area as possible. It would be collected by a simple pool beneath and returned to the system. Such a system would be of particular advantage in the lunar environment if low mass, low vapor pressure fluids could be obtained from indigenous materials. Droplet control and aiming would no longer be as critical as in the space environment; however, the system would need to be shaded from the Sun when it is in operation. While this system is far less developed than the systems previously discussed, its promise is so high that it warrants serious consideration for future use, particularly in response to our growing needs for improved power management.

Belt radiator concepts: The belt radiator concept is a modification of the liquid droplet concept in which an ultrathin solid surface is coated with a very low vapor pressure liquid (see fig. 39). While the surface-to-volume ratio is not limited in the same fashion as for a cylindrical heat pipe, it does not quite match that of the liquid droplet radiator. However, this system avoids the problem of droplet capture by carrying the liquid along a continuous belt by surface tension. The liquid plays a double role in this system by acting not only as the radiator but also as the thermal contact which picks up the heat directly from a heat transfer drum. Variations on this scheme, in which the belt is replaced by a thin rotating disk, are also feasible but have yet to be fully assessed.

Collected in Space Resources NASA SP-509 vol 2

You noted that having too many radiators distributed about an axis causes them to radiate into each other. It all boils down to what's known as a face factor, essentially how much of the radiation released by a surface is intercepted by another one. For two plates of equal length separated by an angle alpha (α), the face factor is:

F = 1 - sin(α/2)

So you can see right off the bat that for 2 radiators opposite each other, α = 180°, α/2 = 90° and the face factor is 0, no interception. But go up:

Multiple Radiator Panels
  • # number of radiators spaced around the ship's long axis
  • Face Factor how much heat radiation from a radiator is wastefully intercepted by another radiator
  • Emit how much heat is effectively radiated by the total radiator array, in units of single radiator panels
  • Efficiency how efficient is this array at getting rid of heat, single panel = 1.00 or 100%

The second column is face factor, the third column is how much is emitted relative to a single surface, and the last is "efficiency", how much every individual panel is emitting relative to a single unlimited surface (efficiency is just 1 - F = sin(α/2), which is itself the face factor for the surface relative to its unobstructed surroundings). As you can see, it falls off very very quickly; the third radiator is only 60% as effective (goes up from 2.00 to 2.60), and the fourth adds to this only marginally (goes up from 2.60 to 2.80). Unless there really isn't room to simply stretch out the panels, it just doesn't seem worth it to pack more than 3 about an axis, and even 3 might be a stretch.

Neat thing is, using the face factor you can figure out the efficiency of radiators in weird geometries. My textbook has face factors for cylinders, enclosed spaces and plates of unequal size, if you so desire, which is to say I could tell you how much is going into your spaceship, or out an open dock.

Another neat side effect of face factors is that you can make a radiator more efficient per given mass by poking holes right through it, since the inner surface of the hole radiates at least partly out into its surroundings (the rest radiates back into itself, but that isn't really a problem). This reduces efficiency per unit area (though interestingly not by much for giant holes), and the panel is significantly weaker as a result (even more than you'd think, since the hole provides an area of stress concentration — it can reach multiples what it would normally there), but for very small holes that are very close together, you can get efficiencies per mass that are many times higher than they would be for a straight panel.

Take this arrangement: a square grid of side L, with a hole in the center of each square and one on each vertex, each hole being of radius 0.35L so that the holes at the vertices are nearly touching that in the middle. Say, for L varying from 0.1 to 1, 10 and 100 times thickness, relative mass efficiency goes up 4.342, 4.010, 1.828 and 1.094 times(by the way, because the relative size of the hole is the same, you need 4.660 times the area of panelling to get the same mass as a continuous radiator).

(ed note: the above was orginally erroneously writen as 1.199, 2.204, 4.184 and 4.609 times)

You can force even more holes into there if they're arranged hexagonally; take a hexagon of side L, with a hole at the center and one each vertex, you can reach a radius of up to 0.5L. Now, for L varying again from 0.1 to 1, 10 and 100 times the thickness, relative mass efficiency goes up 10.741, 9.610, 2.763 and 1.193 times over (in this case, you need 10.74 times the area of panelling to get the same mass as a continuous radiator)!

(ed note: the above was orginally erroneously writen as 1.486, 5.035, 9.816 and 10.644 times)

Given that "every gram counts", it's almost certainly worth the fragility. You could probably thicken the panels somewhat to make up for it and hit a sweetspot

From Zach Hajj (aka Zerraspace) (2015)

For one, I hoped to clarify some things in my former post, so I've provided images for the hole arrangements to make them easier to understand.

It turns out I made a little slip with the calculation — it's smaller holes that are close together that give higher efficiency, not large holes. It just means correcting a couple of sentences there — "very small" rather than "very large" holes, and for the figures, for a square grid: for L varying from 0.1 to 1, 10 and 100 times thickness, relative mass efficiency goes up 4.342, 4.010, 1.828 and 1.094 times, and for a hexagonal grid: for L varying from 0.1 to 1, 10 and 100 times thickness, relative mass efficiency goes up 10.741, 9.610, 2.763 and 1.193 times.

(ed note: I tried to make the changes specified in the above paragraph in the prior quote "Multiple Radiator Panels". They are marked in bold red letters.)

Now, for the new stuff. I’ve found out there’s a hard limit to the emission you can get by stuffing more panels about an axis. You see, mathematically, sine can be expressed by a Taylor series, so broken down:

E = n sin(a/2)
= n sin(2π/2n)
= n sin(π/n)
= n ((π/n) - (π/n)3/3! + (π/n)5/5! - (π/n)7/7!+…))
= π (1-(π/n)2/3! + (π/n)4/5! - (π/n)6/7!+…)

As n grows, π/n shrinks, till all terms but the first disappear, leaving us with π! You can think about this in geometric terms too — emission is basically the perimeter of the shape formed by the joining of the tips of each plate, so as n gets really huge, the shape transforms into a circle (in fact, this is how computers build circles, as n-gons with huge values of n). Instead of working to pump out every last drop possible within a confined space by throwing in essentially useless panels, you could replace them with one large cylindrical panel, giving you π emission, the maximum possible, but with 100% efficiency because there’s no self-interception! That being said, you’ll still lose out to two panels on opposite sides of the ship, seeing as those can get double the emission by radiating from both sides.

But don’t discount those extra panels yet. See, the formula I gave you makes one critical assumption — that the length of the plate so greatly outstrips width it might as well be infinite. This is not necessarily the case, so I decided to go and find a more expressive formula, here - . It involves a very complex integral, so using a MATLAB code, I managed to solve it numerically, cross-referencing with the values given on the website to ensure accuracy. The results are presented in the attached Excel file; the values are only somewhat higher for triangular plates.

It seems that for small aspect ratios (width/length is less than ¼ or so), the value of the face factor is closely approximated by the simplistic formula I gave earlier. However, it falls off dramatically as the aspect ratio climbs, and for very large values (above 50 or so), the other panels barely present any obstruction at all!

The reason for this is that emission isn’t all perpendicular from the radiating surface — that’s simply the direction in which it is most intense. Some is emitted at an angle, so for huge aspect ratios, more can escape through the open sides. This is what makes belt radiators and heat pipes so effective, and lets us get any performance improvement out of poking holes in the panels. It also allows you to experiment with novel designs, as you can now have radiators that point right at each other without adjacent surfaces being rendered completely useless, so long as there’s some distance between them.

For one, you can further reduce face factor by swapping out flat panels. Cylindrical radiators fare spectacularly well: if my code is on the money, for length/radius greater than or equal to 5, face factor for up to 12 such tubes arranged radially is 2.45% or less, essentially negligible, and while I could not find a formula for cones, I can tell you performance will be intermediate between flat plats and cylinders. Attack Vector: Tactical’s radiator spikes would seem to be in the realm of possibility, though I wouldn’t vouch for their exact arrangement given in the illustrations (arranging spikes to the front and rear of one another blocks more and more of their surroundings, which kind of defeats the purpose of using them).

Even directly parallel panels can be worked with, as shown here:

(Images for rectangular and circular panels are from Fundamentals of Mass and Heat Transfer, 6th Edition, by Frank Incropera, David Dewitt, Theodore Bergman and Adrienne Lavine; the last is from the online Catalog of Radiation Heat Transfer Configuration Factors, by John R. Howell at the University of Texas in Austin, available here: )

Take two square panels of side X, separated by the same distance (X/L=1), then the face factor is only 0.2, not exactly murderous; you get nearly the same value if the panels are triangular (theta = 45, c/b = 1) or circular, with diameter equal to the separating distance (rj/L=0.5, L/ri = 2). If one dimension of a rectangular panel is much shorter than the distance between them, say a fifth of it or less (either X/L or Y/L < 0.2), then it doesn’t matter what the other dimension is, face factor is always below 0.1, and the same can be said for triangles. You could get some interesting geometries out of this — a homage to the TIE fighter, anyone?

Again, this analysis has to be taken with a grain of salt. None of these is a free lunch. You have to transport heat across that distance of paneling without temperature dropping off too much, or you’re only going to get the real heat disposal near the body of the craft, which just so happens to be where most of the interception takes place. This isn’t a problem when your panels are only a couple or even tens of meters across, but it does become a concern when they stretch kilometers away from the ship body. Moreover, they require more mass in the form of additional support and shadow shield coverage (if the ship needs such), and they present an even greater target for weapons’ fire. Whether or not this always pays off is a valid question.

From Zach Hajj (aka Zerraspace) (2015)

The Glow

What color will the radiators glow? A practical one will only glow dull red. You can use the Blackbody Spectrum Viewer to see what temperature corresponds to what color. If it was glowing white hot, the temperature would be around 6000 Kelvin. This would be difficult for a solid radiator, since even diamond melts at 4300 degrees K.

Optimum Radiator Temperature

Here is some scary math about radiators from Dr. Tony Valle and Ray Robinson, along with some interesting conclusion. Remember that according to the radiator equation the hotter temperature the radiator is run at, the more waste heat it can dispose of.

RocketCat sez

Their "interesting conclusion" is that ya don't wanna design your heat radiator to run at 100% efficiency or the blasted thing will be huge, unwieldy, and bloated with penalty mass. Remember every gram counts!

For the sweet spot between maximum efficiency and minimum mass, design the radiator temperature to run between 3/5 and 3/4 of the power plant's hot end. But if you don't wanna take my word for it, feel free to dive into the following scary math.

It is surprising but there is an optimum temperature ratio at which to run a starship heat exchanger (or similar power source) to achieve maximum free power with a minimum of radiator area. The only assumptions necesary are that the power source obeys the laws of thermodynamics and that the starship may only get rid of waste heat by radiating.

Let us assume that we have a heat engine as a power source with a relative efficiency of η, and an absolute efficiency is η times the Carnot efficiency ε. We can write the available free power, F, as:

F = Qηε = Qη(1 - T1/T2)

where Q is the rate of heat flow into the exchanger and T1 and T2 are the temperatures of the cold and hot sides of the engine, respectively. The waste heat, H, released into the starship is Q - F, or:

H = Q(1 - η + ηT1/T2)
H = F (1 - η + ηT1/T2)/η(1 - T1/T2)

To simplify, we will measure temperature in units of T2 and let T1 be called just T. After dividing through by η the amount of waste heat associated with a given free power F is then:

H = F (η-1 - 1 + T) / (1 -T)

Now this waste heat must be radiated away from the ship. The power radiated by a black body at temperature T and with area A is given by the Stephan-Boltzmann Law:

P = AσT4

with σ a constant depending on the choice of units. Setting these equal to each other gives:

A = F (η-1 - 1 + T) / σ(T4 - T5)

Now we can ask what value of T will give the minimum radiator area. Taking the derivative of A with respect to T and setting it equal to zero gives:

(T4 - T5) - (4T3 - 5T4)(η-1 - 1 + T) = 0

Or, dividing by T3 and expanding:

T - T2 - 4η-1 - 4 + 4T + 5Tη-1 - 5T + 5T2 = 0

After collecting terms, we have:

4T2 + (5η-1 - 8)T + 4(η-1 - 1) = 0

or, dividing through by 4:

T2 + (5/4η-1 - 2)T + (η-1 - 1) = 0

We write η-1 as γ then the solution to the above quadratic can be written:

T = 1 - 5/8γ + 1/8 sqrt(25γ2 - 16γ)

In the special case where the exchanger runs at maximum theoretical efficiency, η = γ = 1 and the equation above gives T = 3/4. This means that the cold side of the heat engine is at 75% of the temperature of the hot side.

This is horribly inefficient as a Carnot heat engine goes, but if the radiator temperature drops, the surface area (and thus mass) must increase, because of the T4 behavior of the radiation law — colder radiators dump heat much less efficiently. This function is fairly flat — as η drops from 1 to a more plausible 0.1, T changes from 3/4 to 4/5

Dr. Tony Valle and Ray Robinson's commentary in Attack Vector: Tactical, "Science Behind The Rules: Power and Heat Generation" (2004)

Take a classic space opera warship. Onboard power is generated by one or more fusion reactors. If the overall power is 2 gigawatts, and the efficiency is 90% (a pretty generous estimate, since projections I've seen for MHD power generation are around 60%) then at full power, the reactors create 200MW of waste heat. At these sorts of power levels the waste heat of the crew, computers, coffee makers, etc. can be ignored. If there are energy weapons, assume they too are 90% efficient and use 500MW of power when fired, generating another 50MW waste heat. Lump in a lot of other minor systems and you get something like 300MW total waste heat that has to be gotten rid of at peak.

Where it gets complex, AFAIK, is the question of how hot you can allow the ship's interior to get. Let's assume that the environmental areas have heat pumps that allow them to stay a fair amount cooler than the engineering areas (since they're not generating the majority of the heat to begin with) and there are no low-temperature superconductors and so forth to worry about. If the engines and weapons can operate happily at 150 degrees Celsius, that's 423 degrees Kelvin. So that's our starting point — the coolant (probably liquid sodium or lithium at that temp) gets that hot before it's pumped through the radiators to cool off again, at which point:

Heat lost [watts] = area [m2] * emissivity * 5.67e-8[Stefan-Boltzmann constant] * T4 [degrees Kelvin].

(ed note: Stefan-Boltzmann law again)

If the radiators are perfectly black (emissivity of 1), and the coolant temperature is 423 degrees K, then in order to radiate away 275MW of heat, the radiator needs be about 150,000 square meters in area (of course it's double-sided, so the actual fin(s) only need to be 75,000 m2). That's a square 275 meters on a side, or roughly a large city block, simply to deal with the ship's own waste heat at full power. If the ship needs to radiate away additional heat due to taking in, say, 400MW of energy from an enemy ship's lasers, you'd probably have to double or triple that figure (and make darn sure to keep your fins edge-on to the enemy ship firing at you! :). Of course all this is very crude and assumes perfect efficiency of a number of things (some of which I'm probably unaware of :). In reality you might get 80% of that theoretical performance. Or perhaps less. And the first thing damaged in a battle would probably be the radiators (big, hard to protect).

The structural mass of a large radiator fin could be a substantial fraction of the entire ship's mass, and that slows down the acceleration of the ship, which needs more power for thrust, which gives off more waste heat, and so on... So the idea of using spray wands and droplet coolants is attractive.

OTOH, if you need to keep the whole ship at a comfy temperature like 20C, then it's almost hopeless. The radiating area required is so enormous that high acceleration isn't practical at all (something like half a million m2).

Another alternative is to design the ship to only radiate away normal, routine power levels, and to boil off propellant to deal with peak loads. But that goes through a lot of propellant pretty fast at high power levels. Dreadnaughts become like modern jet fighters — only good for a few minutes of intense combat before the fuel runs out. Once it's gone, you can't crash, but you have to surrender or be boiled...

(ed note: He is assuming that the radiator temperature is 423K. Some of the other estimates were for radiator temperatures of 1600K to 3000K, which would drastically lower the radiator surface area to 800 m2, if double sided it would be about 20 meters square.)

Radiator Types

Use the "Life Support" radiator data for life support and other low-waste-heat management. Use all the others for high-waste-heat management, such as fission/fusion reactors and weapons-grade lasers.

In each radiators Specific Area data table will be listed Heat Cap., Mass, and Op. Temp.

Heat Cap.: heat capacity in kWth/m2. This is how many kilowatts of waste heat each square meter of radiator can get rid of. Multiply the surface area of the entire radiator by the heat capacity to find the total amount of heat the radiator array can handle. kWth means "kilowatts of thermal energy" (i.e., waste heat) as opposed to kWe which means "kilowatts of electricity".

Mass: specific area mass of the radiator in kg/m2. This is the mass of each square meter of radiator in kilograms. Multiply the surface area of the entire radiator by the specific area mass to find the total mass of the radiator array.

Op. Temp.: the operating temperature of the radiator. You probably won't need this unless you want to fool around with the Stefan-Boltzmann equation. The higher the operating temperature, the higher the heat capacity. Which means the value listed for the heat capacity is only valid if the radiator operates at this temperature.

Use the "Specific Area" values in the tables to calculate the radiator mass.

  1. Decide how many kilowatts of waste heat the radiator will have to handle (from the engine, the power reactor, the laser cannon, etc.)
  2. Select which radiator type to use, and examine its Specific Area table.
  3. Divide the total waste head in kilowatts by the Heat Cap. entry of the table to get the square meters of radiator area required.
  4. Multiply the radiator area by the Mass entry to get the total mass of the radiator required.

or in other words:

radiatorMass = (wasteHeat / specificAreaHeat) * specificAreaMass


  • radiatorMass = mass of radiator array (kg)
  • wasteHeat = amount of waste heat to dispose of (kWth)
  • specificAreaHeat = Heat Cap. from radiator table (kWth/m2)
  • specificAreaMass = Mass from radiator table (kg/m2)

An Attack Vector: Tactical Medium Range Laser 2 has an input energy of 2 gigawatts (GW) and an efficiency of 12.5%. This means that 0.25 gigawatts become laser beam and 1.75 gigawatts turn into waste heat. About par for the course for lasers.

So the laser needs a radiator array that can deal with 1.75 gigawatts of thermal energy = 1,750,000 kWth. This is Step 1.

Looking at the list, by far the radiator with the largest heat capacity is the Molybdenum/Lithium Heat pipe: 469 kWth/m2. This is Step 2.

Divide the waste heat of the laser by the heat capacity of the Mo/Li Heat pipe and we get 1,750,000 / 469 = 3,731 square meters of radiator. This is Step 3.

The Mo/Li Heat pipe specific area mass is 150 kg/m2. Muliply the radiator area by the Mass entry and we get 3,731 * 150 = 559,650 kilograms = about 560 metric tons.

Which sounds like a lot, except if you used an ETHER Charged Dust radiator you'd need about 580 metric tons of radiator.

Note that Step 3 calculates the radiation surface of the radiator. If the radiator is layered flat on the ship's hull, the radiation surface is the same as the physical radiator size. However, if the radiator is attached edge on so it extends out as a fin or a wing, the physical radiator size will be one-half the radiation surface. This is because you can use both sides of the physical fin as radiator surface. Yes, even a liquid droplet radiator. This might not apply for some of the stranger radiator designs, but details are scarce.

Having said that, things are complicated for liquid drop radiators. The radiation surface is the surface area of the droplets. Figuring out the physical radiator size is compilcated, you can find the equations here. There is also Eric Rozier's online calculator.

Note, in the illustrations from the High Frontier game, it uses very strange game-specific terms. Each "mass unit" is equal to 40 tonnes, each thermometer is one "therm" and represents the radiator dealing with 120 megawatts of thermal waste heat (120,000 kWth). When a specific area value was missing I uesd the therm, mass points, and radiator area on the cards to calculate.

Here is a table of the various radiator types. Their area and mass has been calculated as if they were sized to handle 250 megawatts of waste heat.

The table is sorted by array mass, so the better ones are at the top. At least if you want the lowest mass radiator. If the radiation area was an issue you'd probably prefer a Mo/Li Heat Pipe instead.

The life support radiator was included even though it was not intended to handle waste heat over 100 kilowatts or so.

Radiator for 250,000 kilowatts waste heat
RadiatorSpecific area
(Heat Cap.)
Specific area
Marangoni Flow293.04 kWth/m224.4 kg/m2853 m220,816 kg
Electrostatic Membrane51.3 kWth/m24.275 kg/m24,873 m220,833 kg
Hula-Hoop300 kWth/m233 kg/m2833 m227,500 kg
Buckytube Filament293.03 kWth/m248.839 kg/m2853 m241,667 kg
Curie Point212.75 kWth/m235.459 kg/m21,175 m241,667 kg
Tin Droplet38.49 kWth/m26.4154 kg/m26,495 m241,669 kg
Flux-Pinned Superthermal76 kWth/m217 kg/m23,289 m255,921 kg
Attack Vector: Tactical357 kWth/m2100 kg/m2700 m270,028 kg
Bubble Membrane21.01 kWth/m27.00 kg/m211,899 m283,294 kg
Mo/Li Heat Pipe453.54 kWth/m2151.18 kg/m2551 m283,333 kg
Microtube Array102.6 kWth/m234.2 kg/m22,437 m283,333 kg
ETHER212.75 kWth/m270.92 kg/m21,175 m283,337 kg
Ti/K Heat Pipe150.22 kWth/m2100.14 kg/m21,664 m2166,656 kg
SS/NaK Pumped90.83 kWth/m260.554 kg/m22,752 m2166,669 kg
Salt-Cooled Reflux tube75 kWth/m275 kg/m23,333 m2250,000 kg
Life Support0.19 kWth/m23.1 kg/m21,315,789 m24,078,947 kg

      Design factors

     Using the Stefan Boltzmann equation, we can quickly see that a radiator with better emissivity, higher surface area and higher temperature removes more waste heat. 

     On spaceships, it is important to use the lightest possible components for each task. A spaceship with lighter radiators will accelerate faster and have more deltaV, meaning it can go further and do more for less propellant. 
     If we want a lightweight radiator, we want it to have the highest emissivity. We can accomplish this by using naturally dark materials, such as graphite, or painting over shiny metals with black paint. 
     A larger radiator weighs more. We therefore want the smallest radiators possible. To compensate for lower surface area, we can increase the operating temperature. A small increase in temperature leads to a massive increase in waste heat removed. This means that hot radiators are massively lighter and smaller than cold radiators. 

     Further considerations

     A typical radiator accepts coolant from a hot component. The coolant's component exit temperature is the initial temperature at the radiator. The radiator serves as an interface that radiates away the coolant's heat, leading to a lower radiator exit temperature. The coolant is fed back to the component to complete the waste heat removal cycle.

     Heat only flows from a hot object to a cooler object. A radiator can therefore only operate when the component's temperature is higher than the radiator's coolant exit temperature. For example, if a nuclear reactor operates at 2000K, the radiator must work at 2000K or less. 

     The difference between the entry and exit temperatures in a radiator depends on many factors, but generally we want the largest difference possible. This difference in temperature is especially important for power generation. A large difference means more energy can be extracted from a heat source. It also means that less coolant is needed to cool a component.
     This creates problems with realistic designs.
     A general solution is to use two sets of radiators operating at different temperatures: one low-temperature circuit and one high temperature one. It works fine when your low temperature waste heat is a few kilowatts from life support and avionics. Other solutions have to be found for components that must be kept at low temperatures yet generate megawatts of waste heat, such as lasers.  

     For low temperature high heat components, heat pumps must be used. They can move waste heat against a temperature gradient, allowing, for example, a a 1000K radiator to cool down a 500K component. However, this costs energy. Moving heat from 500K to 1000K costs 1 watt to the pump for every watt moved. A realistic pump will not be 100% efficient and will require more than 1 watt to move a watt of waste heat. 
Pump_power = (Waste_heat * Tc / (Th - Tc)) / Pump_Efficiency
     Pump_power is how many watts the heat pumps consume. Waste_heat is how many watts must be removed from the component. Tc is the component's temperature. Th is the radiator's temperature, both in Kelvins. Pump_efficiency is a coefficient.

     A coolant must generally be kept liquid. This imposes a lower and upper limit to the coolant temperature; any colder and it will freeze and block the pipes, any hotter it boils and stops flowing. Water coolant, for example, can only be used between 273 and 373K. More importantly, it limits the temperature difference that can be obtained from a radiator.
     Large temperature differences require that the coolant spend a long time inside the radiator. This requires larger radiators or long, circuitous paths for the pipes. As the coolant becomes colder, it radiates at lower rates, meaning that the last 10 kelvin drop in temperature can take exponentially more time than the first 10 kelvin reduction. There are strong diminishing returns. 
     There are also structural concerns. Large temperature differences impose thermal stresses. These might be too great to handle. Lightweight, stressed radiators are prone to reacting badly to any sort of battle damage, making radiators a weak-spot for any sort of warship.

     All in all, we must keep in mind that there is a restricted range of temperatures between the hot and cold ends of a radiator, and that its performance cannot simply be obtained by using the Stefan Boltzmann equation on the maximum temperature. We cannot use a simple average either, because the coolant loses heat at a quadratically declining rate as it moves from higher to lower temperatures. 
     Here is an example of 1 kg of sodium at 1000K being cooled by a 0.8 emissivity one-sided 1m^2 radiator panel:

     We can see that it takes 17 seconds for the sodium to cool down from 1000K to close to its melting point of 370K. Any cooler and it'll solidify in the pipes. If we average the radiated watts, we get a value close to 11.46kW. This corresponds to an average radiating temperature of 545K.  
     Finally, a radiator suffers stresses when a spaceship accelerates. Some types of radiator break or disperse under strong accelerations, so the spaceship's performance needs to be considered before selecting a design.

     Solid Radiators

     A straightforward design used today.
     It consists of a slab of metal run through with hollow tubes for a coolant to flow. The waste heat conducts out of the coolant and into the radiator material, which radiates it away from its exposed surfaces.  

     This design has a rather high mass per area and low temperature limits, making it one of the worst performing designs. The maximum temperature is whatever keeps the radiator materials both solid and strong, which is important as many metals rapidly lose strength as they approach their melting point. 
     The coolant must remain liquid throughout the cooling cycle, so this limits the temperature difference that can be achieved. Using metals such as tin or salts such as sodium allows for better temperature differences, but pumping them requires specialized, sometime non-reactive, sometimes power consuming equipment.

     The arrangement of radiators around a spaceship must take into account inter-reflection, which is when one radiator's heat is intercepted and absorbed by another radiator. This reduces their efficiency. Anything more than two radiators per axis absorbs some of the heat of another radiator... at four radiators, only 70% of the heat escapes to space, at eight radiators, the efficiency falls to 38%. 
     NASA has studied solid radiators for use in its Nuclear Electric Propulsion concepts. It has specified 2kg/m^2 area density as a requirement for any thermal management system. The ISS's radiators mass 8 kg per square meter, or 2.75kg/m^2 if we only consider the exposed panels.
     So far, only bare carbon fibre radiators operating at 800-1000K have reached this area density. 

     An alternative design achieves better area density by removing the coolant loops and pumps. The heat pipe has a hot end and a cold end, separated by a vacuum.

     Solid coolant is boiled away and then condensed on the cold end, then re-circulated through capillary action or centrifugal acceleration. This method allows for high operating temperatures and does not require any pumps of moving parts, but high mass per area negates many of its advantages.

     On a warship, radiators are a weakpoint. Bright, exposed and hard to defend, they are easy to hit and once the are damaged, they can render a spaceship unable to function. They can mission-kill a warship without ever having to penetrate any armor. Redundant radiators impose a mass penalty. Covering the radiators in plates of armor massively decreases their thermal conductivity between coolant and exposed surfaces, which in turn reduces their efficiency. 
     Solutions for reducing the vulnerability of radiators include pointing them edge-on to the enemy, moving them to the back of the ship, or using retractable designs.

     If all radiators are retracted, the spaceship must rely on heat sinks for its cooling needs. A megawatt heat source can boil off a ton of water in less than seven minutes, so this will only work over very short time periods. 

     High temperature solid radiators run into issues, such as having to deal with the coolant boiling, or having to contain enormous pressures to keep fluids in a supercritical state. The solution is to use solid blocks of metal instead of coolant. Running these blocks like a train around tracks allows for robust radiators that can handle strong accelerations and temperatures up to the boiling points of the coolant blocks (4000K in some cases, if the tracks are actively cooled). The smaller the blocks, down to the size of pinballs, the faster they cool down and the shorter the track needs to be, leading to mass and area savings. 

     Moving radiators
     One of the biggest reasons why solid radiators are so massive is that they need coolant pipes, pumps and heat exchangers to move waste heat from equipment to exposed surfaces.
     To greatly reduce the area density, we can devise a radiator that does not require bulky coolant loops. Instead, we move the radiator.
     Moving radiators rely on the radiator material itself to move through a heat exchanger, out into space to radiate away the heat, then back in.
     Advantages include simpler construction, less fragile designs, less power consumed and very larger temperature differences between the hot and cold ends. This ends up giving them better kg/m^2 and kW/m^2 ratings. However, there are many more moving parts and the radiating surfaces are only a fraction of the volume the radiators take up. Unless very lightweight materials are used, the support structure will negate the mass advantage of such a radiator.

     A disk-and-drum design has a heat exchanger shaped like a drum, rolling against a radiating disk. The hoola-hoop radiator is a large disk held at the tip by a drum heat exchanger. 

     If the wheel or loop is replaced by a flexible or track-linked belt, it can be made to follow various paths. A 'belt-loop radiator' could bring the radiator closer to the spaceship and reduce the structural strength required to survive accelerations or vibrations.

     A wire-loop configuration uses black carbon filaments as the radiating surface. They are flung out of the heat exchanger and held in place by centripetal force. Using high tensile strength materials allows for extremely lightweight loops.

     Rollers can guide the wires instead of centripetal force, thereby becoming an even lighter version of the belt-radiator. High tensile strength materials would be needed, as this allows the rollers and motors to hold the wires under tension to prevent them from sliding around or tangling.

     A rotating disk radiator is a moving radiator where the central component is a spinning disk. Coolant fluid is sprayed at the hub. The low vapour pressure liquid's surface tension causes it to spread into a thin, even film over the disk. As the disk rotates, centripetal force causes the film to flows as it cools to the collector troughs on the edges. This configuration does away with heavy heat pipes and radiator pumps, but requires the use of very low vapour pressure fluids. The disk can be angled inwards, outwards or canted to deal with spacecraft acceleration.

     Bubble-membrane radiators are a 3D version of the rotating disk radiator. Hot coolant is sprayed against an inflated membrane, causing it to spread out into a thin film that very effectively loses its heat. Spinning the membrane causes the liquid film to pool at the bubble's equator, where it is collected and recycled. 
     Advantages includes allowing the use of high vapor pressure coolants and very light construction. Disadvantages include having to contain high pressure vapours in a container that must remain light and transparent.

     Electric radiators
     The designs mentioned so far use physical structures to hold the radiators in place. This imposes some restrictions, such as having to stay within the temperature limits of the support structures, and larger radiators need heavy support to survive even light accelerations. 
     A solution would be to use magnetic forces to hold the radiators in place. Strong magnetic can replace physical support structures for significant mass savings. 

     Examples of such radiators include the flux-pinned radiator. Magnetic fields hold solid radiator components in place. Thermally conductive ribbons transport heat to the magnetic components.
     However, there are complications. Most metals lose their magnetic properties as they are heated, becoming completely insensitive to magnetic fields above their Curie point. Careful selection of the materials used and control of the temperatures is required. 

     A Curie point radiator operates around the temperature at which metallic dust particles lose their magnetism. Iron, for example, loses its ferromagnetism at 1043K.
     The Curie point radiator uses metal filings or even liquid droplets. They are heated to above the curie point temperature and ejected into space, away from the spacecraft. A magnetic field is in place, but they are not affected by it. Iron can be released at temperatures of up to 3134K and collected at 1043K, but Cobalt has a Curie temperature as high as 1388K, is naturally black and boils at 3400K, making it a better coolant. The small size of the particles or liquid droplets allows several megawatts of waste heat to be radiated away per square meter.

     Once the particles cool below the Curie point, they regain their ferromagnetism. They begin to be affected by the magnetic field and are drawn back to the spaceship to be collected.
     Magnetic radiators are excellent solutions for combat damage - at worst, the enemy will disrupt cooling for a few seconds. However, they consume a lot of power and require heavy equipment to generate strong magnetic fields. Any unexpected acceleration or jolt from the spaceship can disperse all the material held in place by magnetic fields. 
     An alternative electric radiator uses electrostatic forces to hold charged particles in place. One example is the ETHER charged dust radiator. Charged particles follow field lines and execute elliptical orbits between the heat exchanger and the collection point. Similar to a liquid droplet radiator, charged particles can be mechanically dispersed and collected efficiently at the other end by oppositely charged scoops. 

     The advantage of electrostatic radiators is that they consume less power, since creating a strong charge differential is easier than extending a strong magnetic field. The equipment is lighter and is less sensitive to temperature changes, since no superconducting or cryogenic equipment is used, and the charged particles can hold a charge across larger temperature differences than they can maintain their magnetic properties. 
     However, the charge carried by the particles can be nullified by natural solar wind or if they come into contact with a conductor. This means they need a clear, short path between heat exchanger and collection point. 

     Liquid droplet radiators
     Liquid droplet radiators do not use any radiating surfaces - they expose the coolant directly to the vacuum. The resultant droplets have incredible surface area for their mass, allowing for rapid cooling and extremely low area density.

     As the coolant does not need to be physically contained, it can be heated to very high temperatures and still cool down very quickly. There are no thermal stress constraints on liquids, so the temperature change can be as extreme or rapid as desired. They do not have to maintain magnetic properties or hold a charge either. This calculator can gives an approximation of an LDR's performance. At 1300K and using 50 micrometer droplets (a fine mist), area density can be as low as 0.047kg/m^2 with an effective performance of 57MW/m^2. This does not include the mass of the heat exchanger, droplet emitter and collector.

     Solutions have already been devised for issues such as the droplets being blown away by solar wind, colliding and merging into larger droplets or moving at different velocities within the droplet sheet. 
     Vapor pressure is still a concern - hot liquids in vacuum tend to evaporate quickly. Special low-vapor pressure coolants must be used, such as liquid gallium, aluminium or tin up to 1200K, lithium up to 1500K. Salting these liquids with a material such a graphite 'grit' or coating them with black ink is necessary to achieve high emissivity. Nano-fluids might allow even higher temperature liquids to be used. Reaching higher temperatures means accepting high coolant loss rates or enclosing the radiating volume in a membrane that condenses and collects vapors. The membrane has to be transparent at the radiating temperatures.  
      The droplets in a liquid droplet radiator need to be spaced evenly and by distances much larger than the droplet diameter — this is to prevent inter-reflection losses from becoming significant.
     Variations in liquid droplet radiators are mostly around how to contain and direct the coolant flow between ejection and collection points.  
      A rectangular LDR has droplet emitter and collector arms of equal length. The collector arm can be made wider than the emitter to catch droplets deviated out of their path by unexpected movements or errors in droplet formation. It might be possible to move the collector arm above and below the droplet plane to intercept droplets when the spaceship is accelerating, as this would cause the droplet sheet to bend away from the plane. 

     A triangular LDR saves mass by using a small collector dish instead of a long arm. However, it is less able to catch deviating droplets or compensate for spaceship acceleration.

     Some LDR designs dispose of the long arms and membranes and instead just spray the droplets into space. The momentum of the droplets makes them follow trajectories that land them right back at the collectors. A fountain LDR shoots droplets in front of an acceleration spaceship. They are scooped up once cool. This method of dispersing droplets produces the lightest possible designs, but there is a risk of droplet losses.

     It works best on spacecraft that gently accelerate over long periods of time, such as nuclear-electric craft on interplanetary trajectories. A shower LDR disperses droplets in front of the spacecraft and has the collectors simply collect them like a ram-scoop. It has less risk of dispersing the droplets than a fountain LDR but requires a long shower-head. 
     Pressure membranes can be an addition to any liquid droplet radiator. They enclose the volume the droplets traverse. Benefits include re-condensing vapours from too-hot droplets, catching stray droplets, allowing for faster droplet velocity and a greater tolerance for droplet sheet instabilities. However, they must remain transparent to all wavelengths the droplets are radiating at, and hold in the vapour gas pressure. These are competing requirements: low wavelength absorption is done with very thin membranes, while high pressure requires thick membranes.  

     Advanced radiators

     Magnetically pumped and focused LDR:

     Ferrofluids at low temperatures and liquid metal at high temperatures can be used as coolant in liquid droplet radiators. They react to eddy currents and magnetic fields, allowing the coolant to be pumped without any moving parts through magneto-hydrodynamics. 

     Magnetic fields can also be used to recover a droplet sheet. Cyclical fields can push and pull on a group of droplets over distances proportional to the field strength. High strength fields could allow droplet sheets to extend over several dozens of meters before being recovered. They would also allow the LDR to compensate for its vulnerability to droplet sheets being dispersed and lost when the spacecraft accelerates by holding the droplets in place.
     Together, an LDR can become extremely lightweight for the area is covers, as no physical support structure has to span its length. 
     Gas coolants: 
     We have looked at solid and liquids as coolants. Gasses can be used too.
     Gas coolants have been used in nuclear reactors already. Carbon dioxide and helium were selected as they are inert and support higher temperatures than water or sodium coolants. 
     In space, the principal advantage of a gas coolant is that it can operate at much higher temperatures than liquid or solid coolants. The same gas could be run from a nuclear reactor to a radiator's tubes and back. It also allows for inflatable structures for the radiators, which can be much lighter than rigid equivalents.

     However, there are limitations and complications. Hot, pressurized gas can be very chemically reactive. While you can push a gas to 3000K+ temperatures, the walls of the pipes containing the gas must also survive these temperatures. Many of the mass savings that come from running a radiator at high temperatures are lost trying to contain and survive the gas coolant. Pumping gas requires much more power per kg moved than liquids, for example.
     Another difficulty is the very poor heat transfer rate between a heat exchanger and a gas. A hot, low density gas like heated helium might have a thermal conductivity hundreds of times lower than a liquid like molten sodium. This leads to difficulties both at the heat exchange interface and the radiating surface interface. 
     A lot of these issues can be solved by using a two-phase coolant loop, meaning it spends some of its time as a liquid and some of its time as a gas. Up to the heat exchanger, the coolant is in a liquid form. It flows through tubes using simple pumps. The heat exchanger is divided into many smaller tubes to increase the contact area between exchanger and coolant. 
     Past the exchanger, the coolant expands. The pressure drop allows it to boil into a gas. This gas travels through a volume enclosed by a hermetic membrane. Through a combination of expansion decompression and the Stefan-Boltzmann law, the gas quickly cools and condenses on the membrane walls. This forms a thin film in microgravity that can be directed towards collection points, where the liquid is pumped back to the heat exchanger. 
     Dusty Plasma radiator:
     This radiator uses conductive plasma, manipulated by magnetic fields, to move and manipulate dust particles. 

     The dust particles suspended in a plasma behave in fascinating ways, still being discovered by the dusty plasma field of research. Interesting behaviours include self-organising into quasi-crystalline structure, building DNA-strand-like bridges through plasma or collecting into disks with empty centres. This is all due to the self-repelling charges the dust particles gain inside the plasma.

     A better understanding of these behaviours can allow for a radiator to combine every advantageous characteristic: wide range of operating temperatures, very low mass per square meter, easily manipulated by electromagnetic and electrostatic forces, low vulnerability to damage and able to survive strong accelerations. 

     The plasma can be quite cold and still serve to manipulate the dust particles. Low-temperature plasma does is safe to manipulate and is quite transparent to the wavelengths the dust particles will be radiating at, meaning it won't heat up or be blown away by thermal expansion. 
     A simple dusty plasma radiator would have plasma trapped in magnetic loops, like coronal loops. Dust would travel along these plasma tubes. More advanced dusty plasma radiators would spray dust particles into a plasma and have it self-organize into thin planes for maximal radiating surface area. Simply changing the ionization state of the particles by running an electric current through the plasma would allow the dust to clump together and follow magnetic field lines straight back to a collector. 

From TOUGH SF: ALL THE RADIATORS by Matter Beam (2017)

Life Support

Specific Area
Heat Cap.~0.19 kWth/m2
Mass~3.1 kg/m2
Op. Temp.? K

Technically you also need radiators to keep the life-system habitable. Human bodies produce an amazing amount of heat. Even so, the life-system radiator should be small enough to be placed over part of the hull, since life-support waste heat is quite tiny compared to nuclear reactor or gigawatt laser waste heat.

Use this radiator type for life-support and other modest waste heat management. Use the other radiators for gigantic waste-heat producers.

The life-system radiators on the Space Shuttle are inside the cargo bay doors, which is why the doors are always open while the shuttle is in space.

Troy Campbell pointed me at a fascinating NASA report about spacecraft design. In the sample design given in the report, the spacecraft habitat module carried six crew members, and needed life-system heat radiators capable of collecting and rejecting 15 kilowatts of heat (15 kW is the power consumption for all the systems included in the example habitat module). The radiator was one-sided (basically layered over the hull). It required a radiating surface area of 78 m2, had a mass of 243.8 kg, and a volume of 1.742 m3. It used 34.4 kg of propylene glycol/water coolant as a working fluid. In addition to the radiator proper, there was the internal and external plumbing. The Internal Temperature Control System (coldplates, heat exchangers, and plumbing located inside the habitat module) had a mass of 111 kg and a volume of 0.158 m3. The External Temperature Control System had a mass of 131 kg, a volume of 0.129 m3, and consumes 1.109 kilowatts.

What this boils down to is that the described system needs about 96 kilograms and 0.405 cubic meters of temperature regulating equipment per crew person. That's the total of the external radiator on the hull and the internal temperature control system.

Simple math tells me the radiator has a density of about 140 kg/m3, a specific area of 3.1 kg/m2 and needs a radiating surface area of about 5.2 m2 per per kilowatt of heat handled (1/5.2 = 0.19 kWth/m2). The entire system requires about 35 kg per kilowatt of heat handled, and 0.13 m3 per kilowatt of heat. But treat these numbers with suspicion, I am making the assumption that these things scale linearly.

Liquid Droplet

Liquid Droplet Radiators use sprays of hot droplets instead of tubes filled with hot liquid in the radiator. This drastically reduces the mass of the radiator, which is always a good thing. A NASA report suggested that for 200 kW worth of waste heat you'd need a 3,500 kg heat pipe radiator, but you could manage the same thermal load with a smaller 500 kg liquid droplet radiator.

The droplet generator typically has 100,000 to 1,000,000 orifices with diameters of 50 to 20 μm. They are a bit more susceptible to damage than the components of more conventional radiators.

A drawback is that the spray is in free fall. This means if the radiator is operating and the ship starts accelerating, the spray will start missing the collector and precious radiator working fluid will be lost into space. Brookhaven National Laboratory has patented a way to magnetically focus the droplet stream. Using a large radiator it will allow the spacecraft to maneuver at acceleration of up to 0.001 g (0.00981 m/s2) which is barely an improvement. The acceleration can be increased but only if the single radiator is replaced by numerous smaller radiators. Which of course makes the sum of the radiators have a larger mass than the single large radiator. Oh, and Brookhaven's patent expired in 1994.

Many liquid droplet designs are well suited for warships, since they do not utilze large fragile panels vulnerable to hostile weapons fire. If a rail gun round or laser bolt passes through a spray of working fluid, it will just make a bit of fluid miss the collector. If weapons fire passes through a conventional panel it will wreck it.

temperature rangecoolant typeexample
250 K – 350 Ksilicone oils
370 K – 650 Kliquid metal eutectics
500 K – 1000 Kliquid tin

Recently, +Matter Beam made an attempt to figure out how to compute the effect of droplet inter-reflection and -absorption in droplet radiators, concluding that the effect was generally small, especially for planar configurations. I was intrigued by the problem, and set out to solve it more accurately.

First, this is a difficult problem, and I had to completely start over at least twice while solving it, because of multiple screw-ups. But once you discover it, the correct math describing it is actually fairly simple. If you have trouble with the subscripts, Chrome on desktop seems to work okay.

Derivation and Listing of Formulae

There are two key effects here: inter-absorption and inter-reflection. In the first, light is absorbed, converted to heat, and re-radiated as thermal radiation. In the second, the light just reflects directly. Both are crucial, but the problem is complicated by the fact that when absorption happens, the energy gets routed through the Stefan–Boltzmann Law, which introduces a fourth power of temperature into an otherwise-simple geometric sum.

Happily, the problem has some neat symmetries, which we can exploit to work around the summation issue altogether.

First, we relate the thermal power coming out of a droplet (Φo, "o" for "out") to a radiometric quantity called "radiance" (Lo). Power is equal to integrating radiance over the hemisphere and then the sphere. You then solve that relation for radiance. I'll spare you the details and give you the answer ("r" is droplet radius):

    Lo = ——————
         4 π2 r4

By definition, this must be equal to the sum of emitted light (Le, "e" for "emitted") and reflected light (Lr, "r" for "reflected") from other droplets:

    Lo = Le + Lr

The emitted light comes from the Stefan–Boltzmann Law (ε is the absorptivity/emissivity of the material, σ is a constant, "T" is temperature in degrees Kelvin). Note we're re-using the relationship between power Φ and radiance L given in the first equation (which holds for any coupled values of power and radiance on a sphere):

    Φe = ε σ T4

         ε σ T4
    Le = ——————
         4 π2 r4

Now, the reflected light is just the portion of the incoming light (Li) that is reflected:

   Lr = (1-ε) Li

The key insight is that, while our droplet might radiate into another droplet, that other droplet is radiating back. Because every droplet is "average", both droplets have the same temperature, radiance, etc. In particular, the incoming radiance from an occluding droplet is the same as the outgoing radiance our droplet is sending back.

Call the fraction of occluded directions "f". In "f" of the directions, our droplet is occluded by another droplet emitting Lo. In "(1-f)" of the directions, we see "0" (assuming the radiance of space is zero, but you could calculate a tiny value from the CMBR if you want). Therefore, the incoming radiance to our droplet is on average:

    Li = Lo*f + 0*(1-f)

Substituting and using the relation between radiance and power on a sphere (first equation):

    Lo = Le + (1-ε) Lo f

        Φo      ε σ T4              Φo   
    ————————— = ———————— + (1-ε) ———————— f
     4 π2 r4    4 π2 r4          4 π2 r4

    Φo - (1-ε) f Φo = ε σ T4

            ε σ T4
    Φo = ————————————
         1 - (1-ε) f

The net power (Φn, "n" for "net") of the droplet can now be computed (remember that "Li = Lo f", as above):

    Φn = Φi - Φo
       = Φo f - Φo
       = Φo (f - 1)
            ε σ T4
       = ———————————— (f - 1)
         1 - (1-ε) f

Thermal energy (J) is related to temperature by the specific heat capacity (c) of the droplet's material. This is considered to be constant (usually accurate as long as no phase change occurs). The relation is simply:

    J = c T

We can combine this with the previous formula, since dJn/dt=Φn and integrate to get functions of time. Thus, we get the energy (or temperature) per time (and obviously we now need J(0) or T(0), the initial energy or temperature of each droplet):

    J(t) =     _______________________________________
              /  3 ε σ       1 - f              1
            3/   —————— * ———————————— * t + ————————
            √      c4     1 - (1-ε) f        (J(0))3

    T(t) =     ________________________________________
              /  3 ε σ       1 - f                1
            3/   —————— * ———————————— * t + ———————————
            √      c7     1 - (1-ε) f        (c T(0))3

Differentiating either of these will give you the power Φn as a function of time. The differentiation is trivial, so I'm not going to write it.

We can also compute the relative efficiency by dividing the equation for Φn when "f>0" by the same equation when "f=0" (the "ideal" case). The result is:

                     1 - f
    Efficiency = ————————————
                 1 - (1-ε) f

Hopefully if I left anything interesting out, you should be able to derive it from one of the given formulae. The "f" factor (again, the fraction of occluded directions) needs to be user-provided, probably as the result of a simulation since I don't have any great ideas on how to compute a reasonable value in closed-form.

Some Examples

By my complete estimate, < 1% occlusion and emissivity 0.4–0.8 is the most plausible configuration. I'm not even sure 100% is possible without ridiculously packed droplets (and it's certainly not possible at all in planar arrangements).

0.001% occlusion, emissivity 0.8 => efficiency ~100.00%
0.01% occlusion, emissivity 0.8 => efficiency ~99.99%
0.1% occlusion, emissivity 0.8 => efficiency ~99.92%
1% occlusion, emissivity 0.8 => efficiency ~99.20%
10% occlusion, emissivity 0.8 => efficiency ~91.84%
50% occlusion, emissivity 0.8 => efficiency ~55.56%
90% occlusion, emissivity 0.8 => efficiency ~12.20%
100% occlusion, emissivity 0.8 => efficiency ~0.00%

0.001% occlusion, emissivity 0.4 => efficiency ~100.00%
0.01% occlusion, emissivity 0.4 => efficiency ~100.00%
0.1% occlusion, emissivity 0.4 => efficiency ~99.60%
1% occlusion, emissivity 0.4 => efficiency ~99.60%
10% occlusion, emissivity 0.4 => efficiency ~95.74%
50% occlusion, emissivity 0.4 => efficiency ~71.43%
90% occlusion, emissivity 0.4 => efficiency ~21.74%
100% occlusion, emissivity 0.4 => efficiency ~0.00%

Some Limitations

The main assumptions used in this analysis are that:

1: All droplets are Lambertian emitters (which means their radiance ("brightness") is the same in every direction). Most light sources and thermal emitters are nearly Lambertian.

2: All droplets are Lambertian reflectors (which means they reflect radiance equally in all directions). This is wrong, especially for metals. Using a more physically plausible BRDF would probably reduce inter-reflection effects even further in planar configurations (that is, decrease the attenuating effects considered here even more).

3: "f" is homogeneous. This is essentially true for the interior region of a radiator. Droplets near the edges have a lower "f" factor, so they radiate more efficiently (again, this decreases the attenuating effects).

Overall, these assumptions are extremely reasonable, and the only way to relax them further would be to do a full numerical simulation for the particular configuration of your radiator. The "right" way, in some sense, is to just simulate it with a path tracer. Unfortunately, I am currently rewriting my path tracer, so I can't show you what this would look like.

Rectangular LDR

Rectangular LDRs have collectors the same width at the droplet generator. The droplet density remains constant across the flight path. It is a simpler more robust design than a Triangular LDR, and has a larger radiating surface (twice the surface area). However the Triangular LDR is lighter (40% less massive) due to its smaller collector.

Triangular LDR

Triangular LDRs have a tiny collector a fraction of the width of the droplet generator. The droplet density increases across the flight path. It is 40% less massive compared to a comparable Rectangular LDR due to the smaller collector. However it is a more complicated design with more failure points, and it has only half the surface area of a same sized Rectangular LDR.

For reasons that have not been made clear to me, Triangular LDR is currently the focus of much of the research and development. NASA likes them better than Rectangular LDRs.

Eric Rozier has an online calculator for droplet radiators here, and for coolant systems in general here. He had this analysis:

Given that the main thing we want to determine is the surface area of the lithium droplets to calculate the heat it can radiate, I decided to build a model of the surface area.

Since no such radiator has been built we have to work with some plausible model data. To model the lithium drops themselves I dug into some meteorological data and found that raindrops typically range in size from 1mm to 3mm, sounds pretty reasonable. Assuming droplets are spherical (a reasonable assumption in zero gravity) then the surface area of any given droplet is of course 4*π*r2.

Working off the wedge based idea you cited here. We then model the full radiating body of the droplets as a triangle, reducing the emitter to a point source for simplification. I'm not sure how space out the droplets should be, but I figure if the distance between any two droplets is roughly twice the radius, the model is probably pretty conservative. Thus for an emitter with distance h from the emitter to the collector, and a collector plate of length h, we get the number of droplets suspended between them to be:

(0.5 * b * h)/(16r2)

We can then model the surface area of the lithium droplets as:

(0.5 * b * h)/(16r2) * 4*π*r2

If you want to modify the spacing of the drops, you can change the inter-droplet gap to q instead of r, rendering the following equation:

(0.5 * b * h)/(4r2 + 4r*q + q2) * 4*π*r2

Eric Rozier

So the equations are:

a = (0.5*b*h) / (16*r2) * 4*π*r2

a = (0.5*b*h) / (4*r2 + 4*r*q + q2) * 4*π*r2


  • a = surface area of lithium droplets in radiator surface
  • b = length of base of radiator triangle
  • h = length of height of radiator triangle
  • r = radius of indiviual droplet
  • q = inter-droplet gap

Liquid-droplet radiators are also a possibility. There do exist liquids which have extremely low vapor pressure at high temperatures — certain organics up to ~600K, liquid metals (esp. lithium) to ~1500K. Using a carefully-designed nozzle to create a fan-shaped spray of fine droplets towards a linear collector results in a very efficient radiator, with minimal weight per unit radiating surface, high temperature, and high throughput.

The radiator would be essentially triangular when "deployed", with the spray nozzle at one vertex and the collector along the opposite side. If the nozzle-vertes is adjacent to the ship body, the collector "arm" will have to extend outwards. Alternately, the collector can be run along the side of the ship, and the spray nozzle extended on a boom and aimed inwards. A series of closely-spaced, narrow-angle nozzles would approximate a rectangular array.

There is always some loss of coolant due to evaporation in vacuum, hence use of liquids with extremely low vapor pressure. You also lose coolant if such a radiator is run under acceleration, unless the collector is over-long and aligned parallel to the thrust axis, which imposes a constraint on system geometry. You also lose coolant if the radiator "panel" is hit by enemy weapons fire; on the other hand there is no mechanical damage unless the much smaller nozzle or collector arms are hit. Bottom line — you'll need a small surplus of coolant, unless you are running a warship, in which case you'll need a large surplus.

If liquid metal is used as the coolant, MHD pumping can be used at the collector arms, resulting in a simplified design with no moving parts. Indeed, in such a case the coolant could also be used as the working fluid in an MHD generator, resulting in a single-fluid, single-cycle power system from primary energy generation to waste heat radiation. Again, a simple, efficient design with no moving parts.

Specific Area
Heat Cap.~38 kWth/m2
Mass6.4 kg/m2
Op. Temp.1030 K

Tin droplet radiator

Atomization increases the surface area with which a fluid can lose heat. A hot working fluid sprayed into space as fine streams of sub-millimeter drops readily loses heat by radiation. The cooled droplets are recaptured and recycled back into the heat exchanger. If tin (Sn) is used as a working fluid, the kilos per power radiated is minimized, using a heat rejection temperature of 1030 K and a total power in the megawatt range (comparable to the game value of heat rejection of 120 MWth per therm). The low emissivity of liquid tin (0.043) is increased by mixing in carbon black, which distributes itself on the surface of the droplet. Evaporation losses are avoided by enclosing the radiator in a 1 μm plastic film, which transmits radiation in the 2 to 20 μm (IR) range. Such a film would continue to perform its function even if repeatedly punctured by micrometeoroids. The illustration shows a triangular liquid droplet geometry. The collector, located at the convergence point of the droplet sheet, employs centrifugal force to capture the droplets. The total specific area is 6.4 kg/m2.

K. Alan White, "Liquid Dropbt Radiator Devebpment Status," Lewis Research Center, 1987

From High Frontier by Philip Eklund
Spiral LDR
Specific Area
Heat Cap.~300 kWth/m2
Mass35 kg/m2
Op. Temp.1200 K

Curie point radiator

A ferromagnetic material heated above its Curie point loses its magnetism. If molten droplets of such a substance are slung into space, they radiate heat and solidify. Once below their Curie temperature, they regain their magnetic properties and can be shepherded by a magnetic field into a collector and returned to the heat exchanger. A 120 MW system operating at 1200 K includes a 13 tonne magnetic heat exchanger and a rotating dust recovery electromagnet on a 25-meter boom, plus 7 tonnes of dust spread in a spiraling disk 27-meters in diameter (35 kg/m2). The usual medium is iron dust, which has a Curie point of 1043 K and is easily scavenged by magnetic beneficiation from regolith.

M.D. Carelli, 1989

From High Frontier by Philip Eklund
Enclosed Disk LDR
Specific Area
Heat Cap.51 kWth/m2
Mass4.3 kg/m2
Op. Temp.1000 K

Electrostatic membrane radiator

This heat-rejection concept, also called a liquid-sheet radiator, encloses radiating liquid within a transparent envelope. It consists of a spinning membrane disk inflated by low gas pressure, with electrostatically-driven coolant circulating on its interior surfaces. The liquid coolant is only 300 μm thick and has an optical emissivity of 0.85 at a temperature of 1000 K. An electric field is used to lower the pressure under the film of coolant, so that leakage through a puncture in the membrane wall is avoided. The membrane has a specific area of 4.3 kg/m2 and 51 kWth/m2.

Shlomo Pfeiffer of Grumman, 1989

From High Frontier by Philip Eklund
Specific Area
Heat Cap.213 kWth/m2
Mass71 kg/m2
Op. Temp.1200 K

ETHER charged dust radiator

To avoid the evaporation losses suffered by radiators that use liquid droplets in space, dust radiators use solid dust particles instead. If the particles are electrostatically charged, as in an electrostatic thermal radiator (ETHER), they are confined by the field lines between a charged generator and its collector. If the spacecraft is charged opposite to the charge on the particles, they execute an elliptical orbit, radiating at 1200 K with a specific area of 71 kg/m2 and 213 kWth/m2. The dust particles are charged to 10-14 coulombs to inhibit neutralization from the solar wind.

Prenger 1982

From High Frontier by Philip Eklund

Heat Pipe

Mo/Li Heat Pipe
Specific Area
Heat Cap.~469 kWth/m2
Mass150 kg/m2
Op. Temp.1450 K

Mo/Li heat pipe radiator

A heat pipe quickly transfers heat from one point to another. Inside the sealed pipe, at the hot interface a two-phase working fluid turns to vapor and the gas naturally flows and condenses on the cold interface. The liquid is moved by capillary action through a wick back to the hot interface to evaporate again and repeat the cycle. For high temperature applications, the working fluid is often lithium, the soft silver-white element that is the lightest known metal. Molybdenum heat pipes containing lithium can operate at the white-hot temperatures of 1450 K, and transfer heat energy at 240,000 kWth/m2, almost four times that of the surface of the sun. The specific area is 150 kg/m2.

David Poston, Institute for Space and Nuclear Power Studies at the University of New Mexico, 2000

From High Frontier by Philip Eklund
Ti/K Heat Pipe
Specific Area
Heat Cap.~153 kWth/m2
Mass100 kg/m2
Op. Temp.1100 K

Ti/K heat pipe radiator

A Rankine evaporation-condensation cycle heat pipe uses metal vapor as the coolant, which is liquefied as it passes through a heat exchanger connected to the radiator. A liquid metal near the liquid/vapor transition is able to radiate heat at a nearly constant temperature. The pipe is made from SiC-reinforced titanium (Ti) or superalloy operating at up to 1100 K, and the working fluid is potassium (K). The pipe is covered with a lightweight thermally-conductive carbon foam, which protects the pipe from space debris and transfers heat to the radiating fins. The total specific area is 100 kg/m2.

From High Frontier by Philip Eklund
Wickless Heat Pipe

Heat pipes are devices to keep critical equipment from overheating. They transfer heat from one point to another through an evaporation-condensation process and are used in everything from cell phones and laptops to air conditioners and spacecraft.

Normally, heat pipes contain porous metal wicks that return liquid to the heated end of the pipe where it evaporates. But engineers are working to develop wickless heat pipes that are lighter and more reliable. Researchers at Rensselaer Polytechnic Institute initiated the Constrained Vapor Bubble (CVB) project to study these wickless heat pipes for use in near-zero gravity environments for aerospace applications.

“Wick structures can be difficult to keep clean or intact over long periods of time. The problem is especially acute for applications, such as NASA’s Journey to Mars mission, that put a premium on reliability and minimal maintenance,” said Professor Joel Plawsky, who heads the Isermann Department of Chemical and Biological Engineering at Rensselaer.

Working with a NASA engineering team, the researchers are conducting CVB experiments at the International Space Station. Plawsky and postdoctoral research fellow Thao Nguyen recently wrote an article about the CVB project in Physics Today, published by the American Institute of Physics (AIP).

“The CVB project is designed to record, for the first time, the complete distribution of vapor and liquid in a heat pipe operating in microgravity. The results could lead to the development of more efficient cooling systems in microelectronics on Earth and in space,” Plawsky said.

A Familiar Technology in an Unfamiliar Environment

A heat pipe is partially filled with a working fluid, such as water, and then sealed. At the heat source, or evaporator, the liquid absorbs heat and vaporizes. The vapor travels along the heat pipe to the condenser, re-liquifies and releases its latent heat, eventually returning to the evaporator, without any moving parts.

In the CVB experiment, Plawsky’s team created a miniature heat pipe, using pentane (an organic liquid) in a glass cuvette with square corners. An electrical resistance heater was attached to the evaporator end. At the other end, a set of thermoelectric coolers kept the condenser temperature fixed. The transparent tube allowed the researchers to study the fluid dynamics in detail, and the sharp corners of the cuvette replaced the job of the wick.

Two main forces affect how a heat pipe performs: capillary and Marangoni forces. The capillary force is what drives the liquid back toward the evaporator. This is the same force that causes liquid to climb up a straw. The Marangoni force arises from a change in the fluid’s surface tension with temperature. This force opposes the capillary force and drives liquid from the evaporator to the condenser.

A Balancing Act

When the amount of liquid evaporating is larger than what can be pumped back by the capillary force, the evaporator end of the heat pipe begins to dry out. This “capillary limit” is the most common performance limitation of a heat pipe.

The researchers expected the same thing to happen in the CVB experiment. But, instead, the evaporator flooded with the liquid. That’s because the Marangoni and capillary forces were no longer fighting against gravity. As a result, the Marangoni force overpowered the capillary force, causing condensation at the evaporator end. However, the net effect was the same as if the heat pipe had dried up.

“As the flooded region grew, the pipe did a poorer job of evaporating liquid, just as would happen if the heater were drying out,” Plawsky said.

The researchers have countered this problem in the next stage of the CVB project by adding a small amount of isohexane to the pentane. Isohexane boils at a higher temperature and has a higher surface tension. This change in surface tension cancels out the temperature-driven Marangoni force, restoring the heat pipe’s performance.

“The School of Engineering at Rensselaer and NASA have had long-standing and productive collaborations on a number of important research projects," said Dean of Engineering Shekhar Garde. “Dr. Plawsky’s heat-pipe research is a great example of our work with NASA to help translate fundamental understanding of liquids into real-world applications here on Earth and in space.”

From WICKLESS HEAT PIPES: NEW DYNAMICS EXPOSED IN A NEAR-WEIGHTLESS ENVIRONMENT by School of Engineering, Rensselaer Polytechnic Institute (2018)

Bubble Membrane

Specific Area
Heat Cap.~21 kWth/m2
Mass7 kg/m2
Op. Temp.800 K

Bubble membrane radiator

This high-temperature concept uses a spinning bubble-shaped membrane to reject waste heat. A two-phase working fluid (hot liquid or gas) is centrifugally pumped and sprayed on the interior surface of the bubble. The fluid wets the inner surface of the sphere and is driven in the form of a liquid film by centrifugal force to the equatorial periphery of the sphere. As the liquid flows along the inner surface of the envelope it loses heat by thermal radiation from the outer surface of the balloon. The use of membranes woven from space-produced carbon nanotubes and cermet fabrics offers a specific area of 7 kg/m2, radiating from one side at 800 K. Liquid metal pumps return the liquid out of the sphere through rotated shaft seals to its source.

Koenig, 1985

From High Frontier by Philip Eklund

Buckytube Filament

Specific Area
Heat Cap.~300 kWth/m2
Mass~100 kg/m2
Op. Temp.1300 K

Buckytube filament radiator

Waste heat may be rejected by moving thousands of loops of thin (1 mm) flexible "Buckytubes" (carbon nanotubes), which radiate their thermal load prior to return to the heat exchanger. Cables constructed of Arm-chair type nanotubes are the strongest cables known, with design tensile strengths about 70% of the theoretical 100 GPa value. The moving filaments are heated by direct contact around a molybdenum drum filled with the heated working fluid, and then extended into space a distance of 70m by rotational inertia. Their speed is varied according to the temperature radiated (from 273 K to 1300 K). The loops are redundantly braided to prevent single point failures from micrometeoroids. Each element is heat treated at 3300 K to increase the thermal conductivity through graphitization to about 2500 W/mK.

Richard J, Flaherty, "Heat-transfer and Weight Aialysis Of a Moving-Belt Radiator System for Waste Rejection in Space", Lewis Research Center, Cleveland, Ohio, 1964.

From High Frontier by Philip Eklund

Flux-Pinned Superthermal

Specific Area
Heat Cap.76 kWth/m2
Mass17 kg/m2
Op. Temp.928 K

Flux-pinned superthermal radiator

Variable configuration radiators take advantage of the surprising physics of high-temperature flux-pinning superconductors. These materials resist being moved within magnetic fields, allowing stable formations of elements. No power or active feedback control is necessary. The radiating elements fly in a flux-pinned formation, not physically touching, but connected by superthermal ribbon. Superthermal compounds hypothetically conduct heat as effortlessly as superconducting materials conduct electricity. The radiating surfaces are graphite foams, which have both a high emissivity (0.9) and a high thermal conductivity (1950 W/m°K) if the heat conducts in a direction parallel to the crystal layers. Operating at 928K, the superthermal radiator has a specific area of 17 kg/m2 and 76 kWth/m2.

Dr. Mason Peck, 2005

From High Frontier by Philip Eklund


Specific Area
Heat Cap.300 kWth/m2
Mass33 kg/m2
Op. Temp.1300 K

Hula-Hoop radiator

By imparting heat to twin washer-shaped disks by direct conduction, the Hula-Hoop radiator avoids the diseconomies of scale that plague fluid radiators. Furthermore, they are robust against micrometeoroid strikes and hostile attack. The two hoop are 100-meters in diameter. They are made of braided cermets coated with graphite, and lubricated in a heat exchanger with tungsten disulfide (WS2). Radiating at 1300 K, each has a specific area of 33 kg/m2 and 300 kWth/m2.

This design is a Philip Eklund original, published here for the frst time.

From High Frontier by Philip Eklund

Marangoni Flow

Specific Area
Heat Cap.~300 kWth/m2
Mass24 kg/m2
Op. Temp.1300 K

Marangoni flow radiator

In zero-g, a surface tension gradient can create a heat pump with no moving parts, or drive micro-refining processes. This phenomena, called Marangoni flow, moves fluid from an area of high surface tension to one of low surface tension. Bubbles operating at 1300 K have a specific area of 24 kg/m2.

G. Harry Stine, "The Third Industrial Revolution," 1979

From High Frontier by Philip Eklund

Microtube Array

Specific Area
Heat Cap.~104 kWth/m2
Mass34 kg/m2
Op. Temp.1000 K

Microtube array radiator

Nanofacturing techniques can fabricate large, parallel arrays of microtubes for high performance radiators. The radiating surface comprises a heavily-oxidized, metal alloy with a 100 nm film of corrosion resistant, refractory platinum alloy deposited on it. The working fluid is hydrogen, which has low pumping losses and the highest specific heat of all materials. This fluid is circulated at 0.1 to 1 MPa through the microtubes, and the heat radiates through the thin (0.2 mm) walls. This allows a specific area of 34 kg/m2, including the hydrogen. The rejection temperature for titanium alloy tubes is from 200 K up to 1000 K, if a high temperature barrier against hydrogen diffusion is used. High speed leak detection capability and isolating valves under independent microprocessor control provide puncture survivability.

F. David Doty, Gregory Hosford and Jonathan B. Spitzmesser, "The Microtube-Strip Heat Exchanger," 1990

From High Frontier by Philip Eklund

Salt-Cooled Reflux

Specific Area
Heat Cap.~75 kWth/m2
Mass75 kg/m2
Op. Temp.1100 K

Salt-cooled reflux tube radiator

In contrast to a heat pipe, that uses capillary action to return the working fluid, a reflux tube uses centrifugal acceleration. This design is more survivable than heat pipes, especially when overwrapped with a high-temperature carbon-carbon composite fabric. Unlike metals, the strength of these composites increases up to temperatures of ~2300K. However, they degrade when subjected to high radiation levels. The working fluid is molten fluoride salts, the only coolant (other than noble gases) compatible with carbon-based materials. Radiating at 1100 K, this radiator has a specific area of 75 kg/m2.

Charles W, Forsberg, Oak Ridge National Laboratory, Proceedings of the Space Nuclear Conference 2005, San Diego, California, June 5-9, 2005.

From High Frontier by Philip Eklund

SS/NaK Pumped Loop

Specific Area
Heat Cap.~93 kWth/m2
Mass61 kg/m2
Op. Temp.970 K

SS/NaK pumped loop radiator

A Rankine evaporation-condensation cycle exchanges heat using a liquid metal as a coolant, which is vaporized as it passes through a heat exchanger connected to the radiator. A liquid metal near the liquid/vapor transition is able to radiate heat at a nearly constant temperature. The usual medium is sodium (Na) or sodium-potassium (NaK), which has a saturation temperature of nearly 1200 K at 1.05 atm. The plumbing is stainless steel (SS) tubes operating at up to 970 K with an emissivity of 0.9. The tube wall is half a millimeter thick to guard against meteoroid-puncture, and each pipe is an independent element so that a single puncture does not cause overall system failure. Molecular beam cameras on long struts scan for meteoroid leaks, which are plugged with pop rivets installed by a tube crawler. Radiating at 970 K from both sides, this radiator has a specific area of 61 kg/m2, including fluid and heat exchanger.

J. Ca/ogeras, NASA/LaRC, 1990.

From High Frontier by Philip Eklund

Attack Vector: Tactical

This fictional radiaor is from the tabletop wargame Attack Vector: Tactical, which is why the description talks about weird units like "power points" and "heat points."

  • One game turn segment is 16 seconds.
  • One power point is 1000 megajoules delivered in 1 segment.
  • So a starship reactor that outputs 1 power point produces at a rate of 1000 MJ / 16 seconds = 62.5 megawatts.
  • 1 heat point is 250 megawatts.
  • 1 hull space holds 20 metric tons.)
Specific Area
Heat Cap.357 kWth/m2
Mass100 kg/m2
Op. Temp.1600 K
(1 - 0.1)

Knowing that our reactors produce 62.5 MW as a base power unit, and using the proof at right, we get an efficiency of 4 J of waste heat per J of power generated. This tells us that wee need to radiate ~250 MW per point of power. The Stefan-Boltzman law states that the surface emits power at a rate of (1-A) * 5.67×108 Wm2 K4 * T4 where A is the albedo, and T the absolute temperature in Kelvins. With an albedo of 0.1, a temperature of 1600K, and 250 MW of output, we need 700 square meters of radiating surface. Extending as a fin, radiating from both sides, this is roughly 18 meters square. At roughly 0.3m thick, and flexible enough to be retracted and extended, we get something that's reasonably 70 tons, or a bit shy of 3 hull spaces. For the sake of game play, one hull space of radiators dissipates 100 MW, or 0.4 heat points.

A civilian (starship) reactor has a built in 16 meter by 16 meter radiator that dissipates its waste heat; this radiator is built into the hull of the ship. This is why civilian reactors require part of the hull of the ship to be unarmored.

Storing the heat before radiation requires a heat sink. A sodium heat sink is ~21.5 cubic meters of sodium, with a density of 0.968 tons per cubic meter. Sodium has a thermal capacity of 28.2 J/mole/K. A mole of sodium weighs 22.98 grams. One gram of sodium absorbs 2.82/22.98 = 1.22 J per K of heat increase. A heat sink of sodium weighs 20.82 tons, raising that mass by 1 K absorbs 25.4 MJ. Sodium melts at 372 K and vaporizes at 1252 K. Pressurized, it remains liquid to 1600 K, our radiator temperature. Assuming a working range of 1300K (room temperature to 1600 K), each heat sink stores 1300 * 25.4 MJ = 33.02 GJ, which is one heat point, assuming other inefficiencies.

Lithium's thermal capacity of 24.8 J/mole/K and molar weight of 6.94 allows 1 gram to absorb 3.57 J per K of heat increase, or 2.92× the heat capacity of sodium. By using 22 tons of lithium, we get triple the capacity of the sodium heat sink.

Water's thermal capacity is 4.186 J/gram/K, 3.426 times that of sodium. Room temperature to boiling is ~85 K, which limits its usefulness. Raising 1 ton of water by 85 K takes 355.88 MJ. One heat point is 33 GJ, and the amount of water needed to store one heat point is 33 GJ/350.88 MJ = 93 tons. Including the extra mass for plumbing, that's 5 hull spaces all told.

Vaporizing water takes 2256 J/g, 6.3× the energy needed to raise it by 85 K. Because the vaporization is not quite perfect, we treat it as 6 heat points removed when the heat sink is vented. The liquid metal heat sinks aren't vented, as vaporized metal would deposit on the sensors of the ship.

From ATTACK VECTOR: TACTICAL Core Rulebook by Burnside, Finley, and Valle (2004)

Discovery XD-1

The spherical pressure hull formed the head of a flimsy, arrow-shaped structure more than a hundred yards long. Discovery, like all vehicles intended for deep space penetration, was too fragile and unstreamlined ever to enter an atmosphere, or to defy the full gravitational field of any planet. She had been assembled in orbit around the Earth, tested on a translunar maiden flight, and finally checked out in orbit above the Moon.

She was a creature of pure space - and she looked it. Immediately behind the pressure hull was grouped a cluster of four large liquid hydrogen tanks - and beyond them, forming a long, slender V, were the radiating fins that dissipated the waste heat of the nuclear reactor. Veined with a delicate tracery of pipes for the cooling fluid, they looked like the wings of some vast dragonfly, and from certain angles gave Discovery a fleeting resemblance to an old-time sailing ship,

At the very end of the V, three hundred feet from the crew-compartment, was the shielded inferno of the reactor, and the complex of focusing electrodes through which emerged the incandescent star-stuff of the plasma drive. This had done its work weeks ago, forcing Discovery out of her parking orbit round the Moon. Now the reactor was merely ticking over as it generated electrical power for the ship's services, and the great radiating fins, that would glow cherry red when Discovery was accelerating under maximum thrust, were dark and cool.

From 2001 A Space Odyssey by Sir Arthur C. Clarke (1969)

The final decision was made on the basis of aesthetics rather than technology; we wanted Discovery to look strange yet plausible, futuristic but not fantastic. Eventually we settled on the plasma drive, though I must confess that there was a little cheating. Any nuclear-powered vehicle must have large radiating surfaces to get rid of the excess heat generated by the reactors — but this would make Discovery look somewhat odd. Our audiences already had enough to puzzle about; we didn’t want them to spend half the picture wondering why spaceships should have wings. So the radiators came off.

From Lost Worlds of 2001 by Sir Arthur C. Clarke (1972)


Typically the percentage of spacecraft dry mass that is propulsion is 3.7% for NASA vessels.

For a list of various spacecraft propulsion systems, go to the engine list.

Habitat Module

This section has been moved here


This section has been moved here

Conserving Payload Mass

Penalty Weight

As you are beginning to discover, mass is limited on a spacecraft. Many Heinlein novels have passengers given strict limits on their combined body+luggage mass. Officials would look disapprovingly at the passenger's waistlines and wonder out loud how they can stand to carry around all that "penalty weight". There are quite a few scenes in various Heinlein novels of the agony of packing for a rocket flight, throwing away stuff left and right in a desperate attempt to get the mass of your luggage below your mass allowance.

Keep in mind that every gram of equipment or supplies takes several grams of propellant. Try to make every gram do double duty.

Tex hauled out his luggage and hefted it. "It's a problem. I've got about fifty pounds here. Do you suppose if I rolled it up real small I could get it down to twenty pounds?"

"An interesting theory," Matt said. "Let's have a look at it -- you've got to eliminate thirty pounds of penalty-weight."

Jarman spread his stuff out on the floor. "Well," Matt said at once, "you don't need all those photographs." He pointed to a dozen large stereos, each weighing a pound or more.

Tex looked horrified. "Leave my harem behind?" He picked up one. "There is the sweetest redhead in the entire Rio Grande Valley." He picked up another. "And Smitty -- I couldn't get along without Smitty. She thinks I'm wonderful."...

...Matt studied the pile. "You know what I'd suggest? Keep that harmonica -- I like harmonica music. Have those photos copied in micro. Feed the rest to the cat."

"That's easy for you to say."

"I've got the same problem." He went to his room. The class had the day free, for the purpose of getting ready to leave Earth. Matt spread his possessions out to look them over. His civilian clothes he would ship home, of course, and his telephone as well, since it was limited by its short range to the neighborhood of an earth-side relay office...

..He called home, spoke with his parents and kid brother, and then put the telephone with things to be shipped. He was scratching his head over what remained when Burke came in. He grinned. "Trying to swallow your penalty-weight?"

"I'll figure it out."

"You don't have to leave that junk behind, you know."


"Ship it up to Terra Station, rent a locker, and store it. Then, when you go on liberty to the Station, you can bring back what you want. Sneak it aboard, if it's that sort of thing." Matt made no comment; Burke went on, "What's the matter, Galahad? Shocked at the notion of running contraband?"

"No. But I don't have a locker at Terra Station."

"Well, if you're too cheap to rent one, you can ship the stuff to mine. You scratch me and I'll scratch you."

"No, thanks." He thought about expressing some things to the Terra Station post office, then discarded the idea -- the rates were too high. He went on sorting. He would keep his camera, but his micro kit would have to go, and his chessmen. Presently he had cut the list to what he hoped was twenty pounds; he took the stuff away to weigh it.

From SPACE CADET by Robert Heinlein (1948)

Long as he had been earthbound he approached packing with a true spaceman's spirit. He knew that his passage would entitle him to only fifty pounds of free lift; he started discarding right and left. Shortly he had two piles, a very small one on his own bed -- indispensable clothing, a few capsules of microfilm, his slide rule, a stylus, and a vreetha, a flutelike Martian instrument which he had not played in a long time as his schoolmates had objected. On his roommate's bed was a much larger pile of discards.

He picked up the vreetha, tried a couple of runs, and put it on the larger pile. Taking a Martian product to Mars was coal to Newcastle.

From BETWEEN PLANETS by Robert Heinlein (1951)

“Is all your cargo aboard? How much did they let you take?”

“A hundred kilos. It’s in the airlock.”

“A hundred kilos?” Norden managed to repress his amazement. The fellow must be emigrating—taking all his family heirlooms with him. Norden had the true astronaut’s horror of surplus mass, and did not doubt that Gibson was carrying a lot of unnecessary rubbish. However, if the Corporation had O.K.‘d it, and the authorised load wasn’t exceeded, he had nothing to complain about.

From THE SANDS OF MARS by Sir Arthur C. Clarke (1951)


In Frank Herbert's DUNE, spacemen had books the size of a thumb-tip, with a tiny magnifying glass.

"If it's economically feasible," Yueh said. "Arrakis has many costly perils." He smoothed his drooping mustache. "Your father will be here soon. Before I go, I've a gift for you, something I came across in packing." He put an object on the table between them-black, oblong, no larger than the end of Paul's thumb.

Paul looked at it. Yueh noted how the boy did not reach for it, and thought: How cautious he is.

"It's a very old Orange Catholic Bible made for space travelers. Not a filmbook, but actually printed on filament paper. It has its own magnifier and electrostatic charge system." He picked it up, demonstrated. "The book is held closed by the charge, which forces against spring-locked covers. You press the edge-thus, and the pages you've selected repel each other and the book opens."

"It's so small."

"But it has eighteen hundred pages. You press the edge-thus, and so . . . and the charge moves ahead one page at a time as you read. Never touch the actual pages with your fingers. The filament tissue is too delicate." He closed the book, handed it to Paul. "Try it."

From DUNE by Frank Herbert

Ruthless Optimization

Other innovations are possible. Perhaps boxes of food where the boxes are edible as well. The corridor floors will probably be metal gratings to save mass (This is the second reason why cadet shipboard uniforms will not have skirts or kilts. Looking up at the ceiling grating will give you a peekaboo up-skirt glimpse of whoever is in the next deck up. No panchira allowed. The first reason is the impossibility of keeping a skirt or kilt in a modest position while in free-fall.) In Lester Del Rey's Step to the Stars all documents, blueprints, and mail are printed on stuff about as thick as tissue paper (have you ever tried to lift a box full of books?).

With regards to low mass floors, the lady known as Akima had an interesting idea:

Unless the deck is also a pressure bulkhead, how about omitting deck plates and beams entirely, and making the floor a metal-mesh version of the trampoline decks used on sailing catamarans? That way, "weights" bearing on the decks would be transmitted into the tubular structure of the hull as an inward tension.

David Chiasson expands upon Akima's idea. There is an outfit called Metal Textiles which produces knitted wire mesh.

The meshes are knitted, as opposed to woven like a screen door. They are manufactured in densities (% metal by volume) from 10% to 70%. There are a wide variety of materials that the mesh can be made from, including aluminum, steels, Teflon, Nylon, even tungsten. Unfortunately, titanium is not on that list, I can only suspect that it must be difficult to get into a wire form suitable for making a knitted mesh.

Direct quote from site's main page: "In compressed form, knitted metal can handle shock loadings up to the yield strength of the material itself. The load may be applied from any direction-up, down or in from all sides."

I can speculate that with some kind of structural forming breakthrough, the mesh could be heated over a (ceramic?) mold to a near-melting point and simply pressed into place, compressing the mesh into a solid.

David Chiasson

Michael Garrels begs to differ:

I need to point out some issues with the idea of mesh floors.

First off there's the idea that bulkheads have to be bulky. In nautical settings, bulkheads have to be bulky to withstand the large pressure of water, to mount things like hatches on, and to provide overall rigidity to the ship during turning and impact. Most partitions in a spaceship would be a thin pressure membrane sandwiched between a mesh to avoid punctures. The skin on the Apollo lander module was thinner than common aluminum foil. If all you're trying to do is partition, pull up pictures of Skylab - you'll see curtains and isogrid all over the place.

Next is your distinction between floors and walls. Unless there is spin or thrust, there will be no such distinction.

Which brings us to the most important point - the floor that you're currently standing on isn't made out of mesh for a reason. Remember that classic description of a gravity well with a weight on a rubber sheet? Many building codes don't limit the weight allowed on floors but instead the amount of deflection allowed. Floors have to be bulky with occasional beams - otherwise you'll never be able to wheel a torpedo or a gurney, and debris will roll toward where you're standing. It might work in a hallway, or as on your boat for stowage of light items, but not for spans more than a couple meters at 1 g using real materials - especially if you want to mount something like a chair and a console in the center of the cabin.

Michael Garrels


If you are dealing with a conventional spacecraft ruled by the iron law of Every Gram Counts, a stowaway is a disaster. If they had not jettisoned a payload mass equal to their mass, there will not be enough propellant to perform the vital maneuvers. The ship will run out, and go sailing off into the Big Dark and a lonely death for everybody on board.

Even if the stowaway jettisions enough mass, there probably won't be enough breathing mix and food aboard for the additional person. Everybody will suffocate and/or starve.

As Dr. Feynman observed about the Challenger disaster, "Nature cannot be fooled." If the equations say that your spacecraft does not have enough fuel, they don't mean "maybe."

For survival's sake, the crew will have little choice but to immediately throw the stowaway out the nearest airlock.

This was highlighted in a famous story called The Cold Equations by Tom Godwin. The story is chilling abet scientifically accurate, but it still caused an uproar when it first came out.

But if the ship is a torchship or uber-powerful faster-than-light starship, things are a little less tense. Since they are not actually threatening the lives of the crew, stowaways will be treated more like their terrestrial counterparts if discovered on a sea-going vessel.


(ed note: Homir Munn has set out in his one-man sports-cruiser hyperspace starship on a desperate mission to fight the Second Foundation. His 14 year old "niece" Arcadia Darrel stows away on board, determined to help.)

     In the luggage compartment, Arcadia found herself, in the first place, aided by experience, and in the second, hampered by the reverse.
     Thus, she met the initial acceleration with equanimity and the more subtle nausea that accompanied the inside-outness of the first jump through hyperspace with stoicism. Both had been experienced on space hops before, and she was tensed for them. She knew also that luggage compartments were included in the ship’s ventilation-system and that they could even be bathed in wall-light. This last, however, she excluded as being too unconscionably unromantic. She remained in the dark, as a conspirator should, breathing very softly, and listening to the little miscellany of noises that surrounded Homir Munn…
     …Yet, eventually, it was the lack of experience that caught up with Arcadia. In the book films and on the videos, the stowaway seemed to have such an infinite capacity for obscurity. Of course, there was always the danger of dislodging something which would fall with a crash, or of sneezing — in videos you were almost sure to sneeze; it was an accepted matter. She knew all this, and was careful. There was also the realization that thirst and hunger might be encountered. For this, she was prepared with ration cans out of the pantry. But yet things remained that the films never mentioned, and it dawned upon Arcadia with a shock that, despite the best intentions in the world, she could stay hidden in the closet for only a limited time.
     And on a one-man sports-cruiser, such as the Unimara, living space consisted, essentially, of a single room, so that there wasn’t even the risky possibility of sneaking out of the compartment while Munn was engaged elsewhere…
     …She tried to poke her eyes outside the door without moving her head and failed. The head followed the eyes.
     Homir Munn was awake, of course — reading in bed, bathed in the soft, unspreading bed light, staring into the darkness with wide eyes, and groping one hand stealthily under the pillow.
     Arcadia’s head moved sharply back of itself. Then, the light went out entirely and Munn’s voice said with shaky sharpness, “I’ve got a blaster, and I’m shooting, by the Galaxy—”
     And Arcadia wailed, “It’s only me. Don’t shoot.”…
     …After a wild moment in which he almost jumped out of bed, but remembered, and instead yanked the sheet up to his shoulders, Munn gargled, “W … wha … what—”
     Arcadia said meekly, “Would you excuse me for a minute? I’ve got to wash my hands.” She knew the geography of the vessel, and slipped away quickly.

From SECOND FOUNDATION by Isaac Asimov (1953)

Stowaways in Space

     When I sat down to write this it was with misgivings. The idea of anyone hiding on board a ship in space seems a little farfetched to me. When Traveller was released in the ’70’s and people began SF roleplaying there was not an inkling that one day everything would have processing capacity and our wallpaper would have micro sensors in it.
     If you wanted to find a stowaway you had to, you know, look for the little sneak yourself.

     Perusing the rules for the anti-hijack program only indicates it will lock people out of control systems. 
     The classic Starship Operators’ Handbook from Digest Group went into great detail about how a ship’s computer could in fact scan a person for weapons, emotional state and then turn snitch on them long before they’d get to anyplace sensitive. However, we must also realize your computer is already pretty heavily tasked and internal sensors and AI cost credits. Not every ship will have them, certainly not many heavily mortgaged tramp fusers plodding along and plying their trade. So let’s go over some ways to stow away.

     Stowing away is defined for my purposes as obtaining passage onboard a ship through nonviolent but illegal means. If you’re holding the captain at gunpoint till he breaks orbit congratulations — you’re a hijacker.
     The easiest way of getting free passage is if you are a powerful telepath (“Aha! Over … wait I was wrong. Just a potted plant in here.”) Let the computer tag you for a drifter all it wants. Who cares if the men sent to grab you don’t seem to see or remember you? The Traveller race, Dronyne, with their cloaking ability would excel at this. S&S Andromedans used to have a molting period where they were invisible and surely an exceptional psi could be found in any race desired. The Dralasites from the Star Frontiers game weren’t invisible but darned near boneless and could easily hide in nooks and spaces no human could reach.
     The problem with all those species is when people begin noticing the missing oxygen, water, and food or see your little footprints on the deck using a thermograph or hear the head flushing. After that it’s a simple matter to order the crew into spacesuits and start evacuating sections of the hull. Note that to get the really satisfying WHOOSH and suck people off their feet we’re talking opening the hangar bay doors.

     Another use for psionics that doesn’t involve making like a Jedi (I am not the stowaway you are looking for …” “Doesn’t matter, you’ll do!”) is to use a power that lets you slow your body’s metabolism down to nothing. Mail yourself wherever in an airtight crate with an alarm clock.
     Medical Fast Drug can serve the same purpose. Make sure you have enough. It also sucks when people do find you as you can’t put up much of a fight (the ship may land before your first punch does.)

     Some stowaways will attempt to ride out their trip in a cargo container keeping a very low profile indeed. These crates are often elaborate affairs with thermal and sonic shielding, cryogenic devices and heat sinks to store and eliminate any thermal traces. This is not too odd when you realize money is not the only reason people stowaway. There are many reasons to let people think you are still on planet. Just be sure you take everything you need. Remember the stowaway who hid in a crated ATV swaddled in chill cans and thermal insulating blankets with water and food bars to spare. They caught him when he emerged after three days to use the bathroom. Yes zero sediment food bars are a thing. Buy some. For added safety bring your own breathing equipment in case they decide to evacuate the cargo bay for whatever reason.
     Of course there are people who convert cargo containers into stowaway modules. Some of them are better than the accommodations you pay for legally.

     There is another way to stow away that few people outside the well travelled circles. We’ve seen how institutions like CT’s Travellers’ Aid Society that gives high passage tickets as dividends.
     Now again back in the ’70’s we figured they were printed tickets (on very nice cardstock) and the recipients lugged them around in wheelbarrows or some such. Actually this works fairly well if you think about it. Give the person a ticket on physical media at each aid station or whatever dated and with a ledger (electronic or physical) that shows the receipt of said tickets. He presents himself, gets a ticket and gets his book stamped all very legal. Note that duplicating a high passage may mean presenting yourself as the recipient or forging a paper trail to prove its sale to you.
     I think those tickets would be very hard to duplicate. Hard but not impossible. counterfeit a high passage and suddenly you are indistinguishable from a paying customer. You can sit around your individual stateroom wondering what the poor stowaways are doing this time of year.
     For my part if I could forge high passage tickets I’d trade them in for the credits then spend them on a middle passage and have a thousand credits to spend on my trip. But then you blur the lines between scam artists and stowaways.

     Another way to blur the lines is to introduce squatters. When you’re hiding aboard a ship you’re a stowaway. When you hide aboard a station you’re a squatter. Being a squatter is an order easier (of course you do need to get to the station first.) Stations are bigger than most ships providing more hiding options. A station with a lot of personnel passing through might not notice the discrepancy in life support and since a station doesn’t go anywhere your mass will not throw it off course. Large space colonies are the ultimate in squatters and may have dozens or hundreds of people living off the grid … in spaaaaace.
     Sadly the results of being found remain the same: possible spacing and most likely arrest and deportation. In the event of deportation they will probably give you a space suit and re-entry kit (used).
     Squatters may even be tolerated on some stations. They provide a steady supply of dayworkers/scabs (or even thugs) if necessary. They could be engaged in all manner of illegal sales and services. If their uses outweighs the oxygen and water loss the station authorities may turn a blind eye to them for years or even decades.

     Of course they can cobble together a stowaway pod. Why do you ask?

From STOWAWAY THE EASY WAY by Rob Garitta (2017)

(ed note: the spacecraft uses some sort of technobabble antigravity device called the "Field Compensation Drive Generator")

The main field went on, and weight ebbed from the Centaurus. There were protesting groans from the ship's hull and structure as the strains redistributed themselves. The curved arms of the landing cradle were carrying no load now; the slightest breath of wind would carry the freighter away into the sky.

Control called from the tower: ‘Your weight now zero: check calibration.’

Saunders looked at his meters. The upthrust of the field would now exactly equal the weight of the ship, and the meter readings should agree with the totals on the loading schedules. In at least one instance this check had revealed the presence of a stowaway on board a spaceship — the gauges were as sensitive as that.

’One million, five hundred and sixty thousand, four hundred and twenty kilograms,’ Saunders read off from the thrust indicators. ‘Pretty good — it checks to within fifteen kilos. The first time I’ve been underweight, though. You could have taken on some more candy for that plump girl friend of yours in Port Lowell, Mitch.’…

…There was no sense of motion, but the Centaurus was now falling up into the summer sky as her weight was not only neutralised but reversed. To the watchers below, she would be a swiftly mounting star, a silver globule climbing through and beyond the clouds. Around her, the blue of the atmosphere was deepening into the eternal darkness of space. Like a bead moving along an invisible wire, the freighter was following the pattern of radio waves that would lead her from world to world.…

…With the silence of limitless power, the ship shook itself free from the last bonds of Earth. To an outside observer, the only sign of the energies it was expending would have been the dull red glow from the radiation fins around the vessel’s equator, as the heat loss from the mass-converters was dissipated into space.…

…An hour after take-off, according to the hallowed ritual, Chambers left the course computer to its own devices and produced the three glasses that lived beneath the chart table. As he drank the traditional toast to Newton, Oberth, and Einstein, Saunders wondered how this little ceremony had originated. Space crews had certainly been doing it for at least sixty years: perhaps it could be traced back to the legendary rocket engineer who made the remark, ’I’ve burned more alcohol in sixty seconds than you've ever sold across this lousy bar.’

Two hours later, the last course correction that the tracking stations on Earth could give them had been fed into the computer. From now on, until Mars came sweeping up ahead, they were on their own. It was a lonely thought, yet a curiously exhilarating one. Saunders savoured it in his mind. There were just the three of them here — and no one else within a million miles.

In the circumstances, the detonation of an atomic bomb could hardly have been more shattering than the modest knock on the cabin door…

…A stowaway was simply impossible. The danger had been so obvious, right from the beginning of commercial space flight, that the most stringent precautions had been taken against it. One of his officers, Saunders knew, would always have been on duty during loading; no one could possibly have crept in unobserved. Then there had been the detailed preflight inspection, carried out by both Mitchell and Chambers. Finally, there was the weight check at the moment before take-off; that was conclusive. No, a stowaway was totally…

(ed note: turns out the stowaway is Henry IX, crown prince of England. He had been trying to travel into space for years but the stodgy prime minister wouldn't hear of it. Henry was smuggled aboard with the aid of the two British crewmen Mitchell and Chambers.)

Saunders swallowed hard. Then, as the pieces of the jigsaw fell into place, he looked first at Mitchell, then at Chambers. Both of his officers stared guilelessly back at him with expressions of ineffable innocence. ‘So that's it,’ he said bitterly. There was no need for any explanations: everything was perfectly clear. It was easy to picture the complicated negotiations, the midnight meetings, the falsification of records, the off-loading of nonessential cargoes that his trusted colleagues had been conducting behind his back. He was sure it was a most interesting story, but he didn't want to hear about it now. He was too busy wondering what the Manual of Space Law would have to say about a situation like this, though he was already gloomily certain that it would be of no use to him at all.

It was too late to turn back, of course: the conspirators wouldn't have made an elementary miscalculation like that. He would just have to make the best of what looked to be the trickiest voyage in his career.

From THIS EARTH OF MAJESTY by Arthur C. Clarke (1955)

Atomic Rockets notices

This week's featured addition is COMPOSITE/CERMET NTR COMPARISON

This week's featured addition is NUCLEAR ORION SPECIFIC IMPULSE

This week's featured addition is SWIMMER STARSHIP

Atomic Rockets

Support Atomic Rockets

Support Atomic Rockets on Patreon