Monday 4 January 2021

Moto-Orion: Mechanized Nuclear Pulse Propulsion

The Orion nuclear pulse propulsion concept has been around for over six decades now. It is powerful and robust, but lacks the flexibility and features we expect from many more modern designs.

Can we give it those additional capabilities?

That cutaway is one of Matthew Paul Cushman’s amazing pieces.

Basic overview of Orion

William Black has plenty of great Orion artwork.
There is a lot of information on Project Orion, available mostly here and here. It is best to read through them to gain a complete understanding of how it works. We’ll only give a simple overview to start.

Project Orion’s design for a nuclear pulsed propulsion system was pretty simple. A physical plate of steel, protected with a thin layer of oil, faced a plasma jet from a nuclear shaped charge. The force of that blast was translated into useful thrust for the Orion spaceship.

In this manner, a propulsion system could tap into the immense power of a nuclear detonation while sidestepping the heat management issues that would normally come from handling such an output. Its thrust was huge, enough to lift thousands of tons into orbit, and so was its efficiency, with an effective Isp of 2,000 to 12,000s. That’s five to thirty times the specific impulse of a chemical rocket, with thrust and efficiency that only gets better as you scale it up. We call this combination of high thrust and high efficiency a ‘torch drive’; a term from ‘Golden Age’ science fiction where authors did not want to spend pages explaining things like deltaV limits and interplanetary trajectories to their readers. A torch drive lets you point at your destination and accelerate to get there. Even today, sci-fi loves this solution.  

It did have drawbacks though. The fissile fuel in each nuclear pulse charge is inefficiently used, with the majority being wasted. This was because each pulse had to be small, so as to not obliterate the pusher plate, and therefore could not produce the better burnup ratios of large nuclear charges. The rate at which these pulses were ignited could not be varied by much either. Timing the pulses with the motion of the pusher plate, so that the blast would meet the suspension system in the right position, was essential.

There were three parts to the suspension system. The first is the pusher plate itself. When struck at a precise angle, it could be accelerated at 50,000g or more without being bent or twisted. It first slams into a gas bag, that acts similar to how a car’s airbags are used in a car crash, to turn a sharp shock into a more gradual shove. Momentum from the plate is then transferred to a set of pistons at a much slower rate. These pistons are connected to rigid springs that convert the series of pushes into a continuous acceleration. When the timing is right, the literally well-oiled machinery is very strong. When the timing is off, things break down.

The suspension cycle, in short

If one charge ignited too early, then only a fraction of the suspension length can be used to absorb the blast’s momentum, so it gets translated into a hard jolt. Ignited too late, and it would further accelerate an already retreating pusher plate, with potentially devastating consequences. A complete misfire isn’t great either. The suspension arms would only be partially compressed, and so would not reach full extension on the rebound and it would become unsafe to receive another nuclear blast. The Orion spaceship would have to wait for the suspension to wobble to a full stop, and then use a half-powered charge to restart it from a fully compressed state. Waiting to restart the suspension cycle isn’t a nice position to be in when launching off a planet. 

Another drawback was the inability to convert any of the nuclear pulse drive’s immense output into electrical power. The two-step suspension system simply acts as a fancy spring to transfer momentum between the nuclear blasts and the spaceship. Most of the time, this is not an issue.

Liftoff from a planet or moon’s surface does not take long, so stored power is sufficient. Cost-efficient interplanetary travel consists of short uses of the main propulsion system followed by long periods of coasting, during which solar panels can be deployed.

An Orion warship accelerating, from the sadly incomplete sequence here.

However, some of the more demanding applications require a lot of onboard power. Military spaceships especially want the ability to both accelerate out of harm’s way, while producing plenty of electrical power to feed lasers, RADARs and other energy-intensive equipment. Fulfilling this requirement means sacrificing payload capacity to mount an onboard nuclear reactor or some other heavy solution. It’s also a problem for very fast transports that want to use the Orion engine as much as possible; they can only extend tiny solar panels while accelerating as anything bigger would get burnt off by the nuclear blasts. 

Of course, there are many other problems too, that we won’t go into more detail this time. The fact that each nuclear charge is a fully functional nuclear warhead, for example, means that a crash-landing would spill out a full nuclear arsenal, worthy of arming a superpower.

Or that the main propulsion system of an Orion ship cannot be used to turn, so huge Reaction Control thrusters would be needed for every single maneuver.

We cannot ignore the existence of more modern and more refined nuclear pulse propulsion designs either. Orion was dreamt up in the 1960s and a lot has happened since then. 

Mini Mag-Orion.

Most notably, Mag-Orion and variants thereof. Instead of a physical pusher plate, a magnetic nozzle is used to capture the momentum of nuclear-generated plasma. Fully self-contained bombs are replaced by subcritical masses of uranium. They have to be detonated by external compression devices, such as a Z-pinch or a magnetic pulse. The result; they are completely safe in storage and gain a not-bomb-like-at-all quality. Generating electrical power is a simple repurposing of coils in a magnetic nozzle into Magnetohydrodynamic generators, and turning is accomplished by unequally deflecting the plasma within the nozzle one way or another. 

However, these more advanced designs cut away at the awesome potential of an Orion drive. The need for large magnets, cooling systems for the nozzle, capacitor banks for the ignition system, all add a lot of weight. Designs of this type have much lower thrust than the original Orion design. They can’t take off from large planets or even operate inside an atmosphere. They move away from that brutal, simple and resilient character that a nuclear Orion engine has, to become something flimsier and more complicated. Perhaps that is an unacceptable compromise, especially for someone seeking specific capabilities, or a sci-fi author aiming for a special aesthetic.  

Ad Astra Game's RocketPunk, seeking that aesthetic.
Could we solve some of the original Orion’s most glaring drawbacks without moving too far away from the image of an atomic piston engine from a bygone era?


We alter the 60 year old design by giving it a crankshaft. It won’t be directly connected to the pusher plate - it can be connected behind the main suspension arms, so that it doesn’t have to receive the shock from a nuclear blast directly and become unreasonably long and heavy as a result. 

The crankshaft is connected to a crank that turns a large wheel. Depending on the pulse rate of the Orion drive, this wheel will turn at 54 to 69 RPM. A gear train would be needed to increase the RPMs into the thousands, suitable for an electric generator. Also necessary is a counter-torque mechanism, such as a second wheel or even just a counterweight turning in the opposite direction. 

Please note that the depiction in the diagram above isn't perfect, as all these mechanisms have to find a place in between the springs, hydraulics and other machinery above the suspension arms. A different arrangement would take up less room, but be harder to read visually.

The concept is similar to a wind turbine and its generator, except the blades are replaced by a nuclear pulse-driven crank.

The power that can be extracted through the crankshaft will be a fraction of the mechanical energy delivered through the Orion drive’s suspension. This is already a small percentage of the nuclear energy released by the pulse charges. The USAF design for a 10m diameter nuclear spaceship has a fantastic 32.9 GW output, but this is only 0.78% of the energy released by 1 kiloton yield blasts every second. We’ll call this the Motorized Orion or Moto-Orion. 

In practice, the electrical power that can be derived from an Orion drive will depend on the mass of the electrical generator and the equipment needed to manage waste heat. A high performance generator would have an efficiency of over 95% and a power density in the tens of kW/kg. Waste heat will be the main obstacle to generating a lot of electrical power, especially as electrical generators tend to operate at lower temperatures. As discussed in a previous post, temperature is the biggest factor in allowing for lightweight heat management systems. 

A generator would typically want to operate at room temperature 300K, but this would mean huge (and heavy) radiators would be needed to handle their waste heat. We want the hottest generators possible. They are mainly limited by the decreased performance of their electrical insulators at higher temperatures. Commercially available motors are available at 570K, but applying research like this could create generators that operate at 770K. However, increased temperatures also increase electrical resistance and therefore cut into the efficiency of a generator. Based on some studies, high temperature efficiency can be held at above 90%. A generator is a motor in reverse, so we will use these same temperatures and efficiencies. 

Estimating the power density of an entire heat management system is quite difficult, but we can make some estimates. 1 m^2 of double-sided 2mm thick carbon fibre radiator fins would be 4 kg and radiate away 8.3 kW of heat at 520K.

Note how this is a slightly lower number than the operating temperature of the generators, as we need a temperature gradient throughout the heat management system to actually move heat from where it is created to where it is radiated away.

With reasonable figures for a silicone oil pump, a microchannel heat exchanger and a +20% margin for assorted pipes, valves and backups, it all averages out to 1.2 kW/kg. 

This seems like a low figure, but it only deals with the <10% of power that becomes waste heat. 1 MW of mechanical energy coming through the crankshaft would become 900 kW of electricity, handled by 45 kg of generators, and 100 kW of waste heat, requiring around 83 kg of cooling equipment. Altogether, this makes for an average power density of 7 kW/kg. This ignores the mass of the crankshaft, counterweight and other mechanisms, but they will be small compared to the rest. There is also the complication of radiator placement; they want to extend out from the hull, but also must stay within the shadow cone of the pusher plate to avoid being disintegrated by nuclear plasma. 

The original USAF 10m Orion had a payload capability of up to 225 tons (on certain missions). If a quarter of this was dedicated just to producing electricity, we could expect it to output 393 MW. That is a respectable amount! 

Here’s what a Moto-Orion derived from that design, with fully scaled radiators, would look like:

Though, it is only 1.2% of the drive power. You could imagine an Orion drive spaceship that extracts more of its output as electricity, but it is fundamentally limited by the difference between the power density of the propulsion system (on the order of 330 kW/kg) and that of the power extracting equipment (<10 kW/kg). Furthermore, equipment that consumes that electrical output will take up an outsized portion of the spaceship’s payload capacity, due to their even lower power density (<1 kW/kg).

There are other ways to generate electrical power.

A linear alternator should be an ideal option. A magnet is simply pushed through a series of conductive coils, producing current as it travels up and down. It is just as efficient as a rotating electric generator, and depending on the exact design used, can operate at the same high temperatures. Even better, it does not produce any sideways torque, is easier to fit in between the suspension arms and is more resilient to vibrations. However, their power density is far lower than that of rotating generators, with 1.49 kW/kg being the best figure mentioned anywhere.

Another option still is to use a high temperature superconducting generator. NASA has designs that aim for 60 kW/kg at the multi-megawatt scale. 

Efficiency is 99%, meaning that 1% of the power becomes waste heat. Thankfully, this heat is produced not in the superconducting magnet, but in the non-superconducting stator. It can reach 570K, so we can use similar heat-management equipment as described above. 1 MW of input power becomes 990 kW of electricity and 10 kW of heat, which are handled respectively by 16.5 kg of generator and 8.3 kg of cooling equipment, for an average power density of 40 kW/kg. 

The downside to using superconducting devices is having to mount the bulky and sensitive equipment needed to keep them in that state. A high-temperature superconductor needs to be kept in liquid nitrogen, which boils at 77K. About 0.01 to 0.1% of the power that a superconducting device handles is expected to become waste heat inside the cryogenic part through ‘AC losses’, where alternating currents create magnetic vortices within a conductor.

Progress is being made into megawatt scale superconducting generators/motors. This Honeywell 1 MW design achieves 8 kW/kg. 
The passive solution to handling this heat load is to just let the liquid nitrogen boil. It can absorb 198 kJ/kg during vaporization, so for every kW a superconducting generator outputs, 5 milligrams per second of liquid nitrogen needs to be expended. 

Using the expendable liquid nitrogen solution, we can have the USAF 10m Orion dedicate 40 tons to electrical production, and 16.25 tons to liquid nitrogen reserves (adding up to a quarter of its 225 ton payload, as before). It would be able to output a whopping 1.6 GW of electricity, but only for 33.5 minutes before liquid nitrogen reserves run out. It’s not too bad; the spaceship would likely run out of pulse charges before it uses up all this coolant. 

The active solution is to use a cryocooler. It raises the temperature of the waste heat to a level where it can be disposed of using radiator panels of reasonable size. If the high temperature superconducting material operates at 100K, then it takes at least 4.7 Watt of cryocooler power to move 1 Watt of waste heat up the temperature gradient to 570K. A realistic cryocooler will achieve 30% of maximum Carnot efficiency, so we increase the power requirement to 15.7 Watts. We choose the 570K temperature target to keep using the cooling equipment from previous calculations (all the better to compare each solution). 

Cryocooler power density for aerospace applications is about 133 W/kg, but 300 W/kg is cited as an achievable goal. Putting these elements together, we have 1 MW of input power becoming up to 1 kW of cryogenic waste heat, which requires 15.7 kW of cryocoolers that mass 52 kg. The active solution brings down average power density to 12.9 kW/kg. It is a respectable figure, better than the non-cryogenic design’s 7 kW/kg, and especially interesting for missions with prolonged engine use with no opportunity to refill on liquid nitrogen..

A USAF 10m Orion that used an actively cooled superconducting generator massing 56.25 tons would produce 725.6 MW as long as the engine is running. 

There is a ‘catch’ to these cryogenic designs though. Superconducting magnets are not known to be resistant to radiation or damage of any kind. It is especially concerning when a nuclear pulse propulsion spaceship bathes itself with penetrating neutrons and high energy gamma rays repeatedly. The magnets cannot be placed too far away from the pusher plate and suspension system either, so they can’t hide in the relatively safe environment the crew enjoys at the other end of the spaceship.


There are two other major benefits to the Moto-Orion. 

The first is during start-up. The original Orion design relied on the suspension system being pre-compressed before the first full-strength nuclear charge could be used. It was the job of a half-strength bomb to get the suspension ready. While this use of fissile material is not too wasteful when compared to the hundreds of bombs that are regularly used, it is very inflexible. Start-up would only be possible a limited number of times, and only when the pusher plate is standing still… not at all comforting when space travel involves must-not-miss burns. It is even worse for a warship that needs multiple successive starts and stops to effect dodges from enemy fire. 

A Moto-Orion can use its electric generator in reverse, to produce torque while consuming energy from battery reserves. It can draw in the suspension arms to a compressed position, or time its pushes and pulls to bring a wobbling plate to standstill more quickly. The batteries can even be charged from another power source, such as solar panels, if battery reserves are depleted. 

This gives the spaceship an unlimited number of restarts. It gains the flexibility to halt and ready its drive at any time. 

The second benefit is recovery after the pulse sequence goes wrong, whether it is late, early or missed completely. 

Accurate suspension cycle for an Orion craft, by ElukkaJ.
A Moto-Orion might be able to react quickly enough to adjust the position of the suspension system in case of a late pulse. Once the nuclear shaped charge moves past its designated ignition point, the spaceship’s motors would draw power to slow down the retreating pusher plate. This could prevent it from being accelerated into the suspension arms at an excessive velocity. 

When things go wrong, unpleasant, up to destructive, g-forces are generated.
An early detonation is especially troublesome. Not only does it erode the pusher plate, it cannot be predicted. The Moto-Orion’s crankshaft and generator can be turned into an additional suspension arm to absorb the unexpected shock, but it would usually be weaker than the massive steel springs the engine habitually relies upon. Still, it can assist in bringing the pusher plate velocity back in line and ready to receive nuclear plasma blasts again. 

When it comes to misfires, Moto-Orion can potentially add velocity to the slower pusher plate (as it did not receive the momentum from the missed pulse) and bring the drive sequence back into correct timing. It can avoid a complete halt by drawing energy from battery reserves, and if it is powerful enough, do so without skipping a beat. 

There are other forms of flexibility, gained indirectly from having access to huge amounts of electrical power. They might not be as flexible in this regard as a nuclear-electric ship could be, as power generation is tied to the use of the engine and not an independent reactor, but many possibilities open up. Orion nuclear spacecraft could deploy drones and beam power to them, by means of microwave emitters or laser beams. They could receive nuclear charges ‘on the fly’ using magnetic scoops. Electrical Reaction Control thrusters can be used, so that the spaceship can turn more efficiently. There are many more possibilities.


An Orion spaceship staging off a aerobraking lander at Mars.
Moto-Orions are safer and more flexible than the original Orions. For a simple transport ship that only uses its engines briefly and wishes to maximize payload, the extra weight is unwelcome. Any craft that carries people might instead find that the additional capabilities and securities are a worthwhile trade-off. Warships would absolutely desire Moto-Orions. The huge amounts of electrical power turn them into terrifying attackers that can both unload with weapons energized by hundreds of megawatts of power while also performing multi-g evasive maneuvers. 

In a science fiction setting, Moto-Orions can deliver the retrofuturistic aesthetic of spacecraft riding on nuclear blasts while also making possible the use of exciting hardware like lasers and coilguns. One setting, RocketPunk, is in development by Ad Astra Games (and by Rick Robinson, who inspired ToughSF). It features Orion-propelled warships battling for Mars in an alternate Cold War future. More engaging action could be made possible with these motorized variants.

The fact that a Moto-Orion connects electrical output with drive power by a single-digit percentage ratio is an interesting feature by itself. We discussed how this avoids troublesome issues such as The Laser Problem, where overpowered lasers have excessive ranges and render maneuvering during ship-to-ship combat useless. Low electrical power and high drive power give room for dynamic combat that is more exciting for readers or viewers. Other types of ‘torch ship’, like a rocket with an immensely powerful fusion reactor, could have better performance than Moto-Orion, but would have proportionally more electrical power - this pushes combat ranges so far out that maneuvering is rendered pointless again. 

The military potential of Orion was always at the forefront.
Another bonus towards dynamic and interesting space combat is an Orion drive’s ability to continuously accelerate and outrun missiles that have less potent propulsion systems. Due to how poorly nuclear pulse propulsion performs when scaled down (burnup ratio and thrust efficiency drop dramatically), a missile would not be able to keep up with a full-sized Orion drive unless it had its own large and expensive pulse propulsion system. They would be excessively expensive, so only smaller and less powerful engines would be available to missiles. Consequently, Orion warships have a good chance of outpacing missiles.

It creates a situation where one side having more missiles than the other does not automatically guarantee a win. Instead, careful use of maneuvers and relative positioning to set up a shot with short-legged missiles is necessary. All the better to read about or play through!

The Project Orion battleship.

We suggest going out and applying these calculations to bring motorized variants to other Orion designs. Huge spacecraft like the 4000 ton USAF 'battleship' could benefit immensely from this concept. 

You could also think about how Medusa could extract electrical power from its tether strokes, or even more outlandish ideas, such as a propulsion system where high velocity kinetic impactors strike a lump of propellant to create a jet of plasma that strikes a pusher plate, like a non-nuclear Orion. 


  1. Fascinating stuff, and one of my favorite blog posts of yours, though I admit as a big fan of nuclear pulse propulsion I'm biased! After reading this I'm sure that my own sci-fi setting, where nuclear pulse propulsion has been commonly used for decades, would employ this variant of the technology. In particular two of the primary applications, easier generation of large amounts of electricity and more easily running the drive continuously for very fast transports, would be very attractive to both civilian and military users. It's exactly what they (and I, as the author of books set in that world!) want.

    I also like how you at least gave Medusa a mention at the end! I was wondering if that would make an appearance! That's a cool variant.

    Lastly, I'd like to say that as usual your prose is engaging and the images are great. Quite enjoyable to read. Keep up the good work! I for one appreciate it!

    1. Thank you Adamas, you've been very supportive.
      Good luck with your sci-fi!

  2. It seems could have sky cities on Venus, and Nuclear Orion could be used for cruise ships to Venus from an Earth orbit. As practical matter, I would say Venus orbit is more significant than Venus sky.
    Though if wanted military fortress, the Venus sky is that fortress.
    If one was weird situation where you had abandon Earth, it seems nuclear Orion are the way to go. Orion work well in an atmosphere. And in Venus atmosphere you have the same deep gravity well of Earth and it doesn't seem like much of problem to set off nuclear bombs in the Venus atmosphere. Or currently most of Venus atmosphere {in terms of mass} can be considered as uninhabitable, it's 94 atm world where only the 2 atm part which can broadly considered useful for humans. You could have some type mining going on at surface and you could have structural stuff {like anchors on ocean floor type thing] down in deep Venus atmosphere, but people aren't going want to live down there- which is mostly quite gloomly and quite dark large part of the time. And in terms mining, obvious thing is the acid clouds- for chemical rocket or on Earth we make a lot sulfuric acid, and sulfuric acid has various uses. So could export the sulfuric acid, if you lower cost of getting things to Venus orbit.
    The sulfuric acid clouds are considered a "greenhouse gas", and if one remove 90% or more of them from the atmosphere, it could considered improvement rather loss of resources {or it's near term resource which can used up}, and centuries later, Venus should cool down a bit. But cooling down doesn't have much effect upon the upper 2 atm of atmosphere, nor does heating it up- which one factor that makes Venus a fortress- a 100 km diameter space rock hitting it, doesn't have much effect, and you can't say that about Earth.

  3. Having studied the concepts on the projectrho page, I've come to the conclusion that if one wishes to miniaturise nuclear pulse propulsion, the best way would be by a combination of the mini-mag Orion and the Gradient Field Exploding Liner:
    Basically, replace the pulsed field/Z-pinch with a static magnetic coil. Superconducting would be best. Use an internal hypervelocity macron accelerator to shoot a subcritical pebble of fission fuel into the field at high speeds, where it gets crushed/implodes. You can also use a small quantity of fission fuel (empty sphere) as a catalyst for igniting an external layer of D-Li. So it would be a Gradient Field Fission-Induced Fusion Drive (because low critmass fuels like americium-252, 251, or curium require reactors to make and are expensive, so it helps if you can use them as a trigger rather than the main power source).
    A SF setting might have reactors powering a naval base (say, in Jupiter orbit) and also breeding high-grade fuels like these to top off the patrolling ships that dock there. And build and startup reactors for friendly colonies, build nukes to use in spacewars, etc
    A less technically advanced military might rely on more easily bred and plentiful U-233. Good enough fuel. And interesting dichotomy would form between a more high end side (Am, Cu with some fusion perhaps) and a more rustic side (using U-233, perhaps in a gun-type device on a regular Orion).

    WRT to the Moto-Orion, would it not be beneficial to engineer a linear alternator INTO the shock absorbers? They're already massive (have to be), so I'd expect the low power density to be less of a problem. And the shock absorbers would already get somewhat hot, so using a LA to turn part of the motion into electricity wouldn't add anything to the heat challenges that wasn't already there.
    I'd also suggest using a coilgun to launch the charges behind the ship, unlike the ammonia-powered gas gun of the original. More control, less space taken up by tanks.

    1. A 'micro-Orion' drive would indeed be very practical. You might want to look into 'PuFF' drive too:

      There would be many complications with engineering combined suspension arms and linear generators. Getting the electricity out of the moving parts would require sliding contacts. The magnets have to withstand vibrations and heavy g forces. You would have no gear ratio to work with if you wanted to use some of the flexibility features (generators as motors) mentioned in the blog post. It wouldn't be impossible... but I just think the vastly superior power density of rotating generators makes these unnecessary.

      Coilgun for launching nukes is a good point!

  4. Why not put nukes in a magnetic nozzle? The efficiency would go through the roof. Just like minimagorion, but without all the z-pinch mininuke fuss. Gigantic nozzles could be made. National stadium in singapore is 310m across. FAST in china is 500m.

    1. The trouble is the strength of the magnetic field needed to contain a full-scale 1kT+ nuclear blast. Your magnetic nozzle would end up being incredibly huge and heavy! Also, that nozzle would need its own complete suspension system, just like regular Orion, because of how much impulse each blast provides.

      It ends up being a worse solution...

    2. Replacing vaporized transmission lines at 1 Hz would be even worse. Minimagorion guys had thought of magorion first, a single 2km coil dragging the ship with cables.

  5. Once again MB, you greatly impress! How would you compare the likely development time of various types of Orion compared to macron-drive ships (fission, fusion, or combo)?
    Do you have any rough development timelines for the many drives and technologies you've included over the years, to *the effect of:
    "At current rates of progress, we can have technology X in 50-75 years, and different technology Y in 100-125 years."?

    Thank you again.

    *Like this on Atomic Rockets:

    1. Thank you!
      All Orion concepts suffer from the fact that they need to be loaded up with hundreds to thousands of self-sufficient nuclear bombs. Even when restricted to space-only use, it is a very concerning vulnerability and a major security issue. Unless a major shift happens in our politics and how we handle nuclear safety, Orion will not happen.

      That's not to say that no nuclear propulsion will materialize. Macron accelerators are an example of nuclear technology that is very hard to weaponize without designing it from the ground up as an offensive tool.

      With all that in mind, I would put the Orion drive at a wildcard, unknowable future, and macron propulsion at the end of the century.

  6. I remember reading in George Dyson's book on Project Orion

    That you could improve fission burn up by using a larger conventional charge for ignition, but the military wasn't interested because it led to a larger bomb. So maybe propulsion with smaller bombs is possible.

    On a side note, how radioactive would the pusher plate get after being used a few times? Would these ships have to keep a large stand off distance from port infrastructure?

    1. A larger bomb would mean that everything else, from pusher plate to suspension system, would have to get proportionally bigger. Not great when you would have had to struggle to put even the smallest Orion into orbit through many conventional launches!

      The total exposure time of the pusher plate to nuclear blasts amount to something like a single second over the course of an entire multi-year mission. That's not enough for it to become radioactive. It will be releasing neutrons for a short while after use, but it is not a permanent change.

  7. Hello Matter Beam, great post! Before this I've always thought that such a system was already thought out and written in some research paper I've never heard of before.

    Though seeing a fleshed out concept really brings some possibilities into the world I'm making.

    Which brings me ask, is there any reason to use other drive systems other than Orion? In the setting I'm making, Orion drives are one of the primary forms of propulsion used in spacecraft. I have been thinking of adding Nuclear Salt Water Rockets (the lithium variety) alongside Orion in order to have that COADE feel to the setting alongside the raw power of Orion, though why anyone would used a NSWR over Orion is still something I'm figuring out.

    Other drives are available, NTR's up to closed cycle gas cores are used for smaller ships in smaller systems like the Jovian's or the ring system of Saturn. Fusion rockets are a thing, though are nothing like a torch drive currently, just having a good balance between delta v and thrust; just because something says 'fusion' on it doesn't mean it's a magic wand, let alone a torch drive.

    Does this make sense? Or should I choose one of the drives (Orion/NSWR) and get rid of the fusion idea go Heinlein all the way?

    Hope your having a nice day!

    1. Thank you!

      The main downsides of Orion are its extreme cost (it uses large amounts of expensive enriched uranium packaged into even more expensive explosive lenses), its limited specific impulse (about 100 km/s unless you redesign it) and its inability to be scaled down.

      Fusion technology is very different from the simple mechanisms involved in fission Orion, so having one or other other is completely unrelated.

      What might be interesting for you is to 'upgrade' Orion to use fusion energy. Each nuclear pulse can be modified from a single stage fission device to a two-stage fission-fusion device, with a big boost in energy and Isp.

    2. Thank you for the reply, it really got me thinking!

      Something that my mind came across when thinking through about using fusion bombs for the bomblets of Orion.

      From my limited understanding of nuclear bombs, the high cost of atom bombs comes from the high amount of weapons grade uranium or plutonium inside the devise. While a fusion bomb of comparable strength is more cost efficient due to smaller amounts fissionable materials inside.

      This would result in cheaper and more numerous bomblets for any given trip along with the increased delta v.

      Am I right? Or is there a flaw in my thinking.

      Also, looking back at the NSWR concept, it's basically an Orion constantly exploding, thus having overall better performance, right?

      Have a nice day!

  8. Another interesting post! I see someone has beaten me to the punch with making the shock absorbers into linear motor/generator sets, which would seem to have the advantage of being somewhat more compact than crank driven generators. I suppose the issue has to be decided by looking at the overall mission: do you want the space inside for generators or do you need it for something else. The question of trade offs is always interesting to contemplate.

    As an aside, using the Orion nuclear pulse drive to turn a crank has a wonderfully steampunk feel to it - it is as if Jules Verne was sitting down with Ted Taylor and Freeman Dyson during the design meetings.

    Several years ago on the Nextbigfuture website, there was an article about using an Orion interceptor against incoming asteroids, which took the entirely opposite track - the proposed vehicle had no shock absorbers at all. I remember some fairly amazing performance figures, like 100G acceleration - this is stripped down kinetic energy impactor after all. Sadly the search function is rather poor and the site has changed greatly so I can't find the link just now.

    1. Hi Thucydides, glad you like the idea.

      I'm under the impression that if the Orion was redesigned from the ground up to use hybrid actuators/linear generators/shock absorbers, it could end up lighter than a crankshaft+rotating generator design.

      A zero-suspension Orion would be a scary thing. Each nuclear blast would hammer it at 50,000g. That's the sort of vibration we use to dig through bedrock...

  9. @ Matter Beam: Thanks again.
    Semi-OT: from your understanding, how similar is Project Starshot's proposed technology to Macron Propulsion? Is it just that they're both "big honkin' space guns" or are there additional similarities?

    @ Unknown: Hey somebody else mentioned Li NSWR! I can find very little about it on the web and it seems to be a good concept (to a layperson). Do you have any additional info on it?

    1. Macron accelerators can be considered to be a form of mass stream propulsion.

  10. @ Keith Halperin
    Well, the source I used comes from Atomic Rockets here:

  11. If the gasoline inside an average car engne is burned and injected at the highest pressure in the combustion chamber, what is this car engine specific impulse?

  12. Thanks again, Unknown.
    I'd seen mention of it here:, and here .
    I wonder where it would go in the AR Drive Table? (Winchell has the usual NSWR in there...)
    I sent info about it to Winchell and Rick "Rocketpunk Manifesto" Robinson just about 3 years ago...
    DISCLAIMER: I'm not an engineer or scientist but I'm curious if the Li NSWR efficiency could be improved with antiproton catalysis?
    Best luck in your world-building....


  13. This comment has been removed by the author.

  14. Easiest way to do propellantless flight is to stage the devices ahead of the craft and have it ramjet the devices to the back of the ship and detonate. This could get you to 0.3c after which you need staged devices. Of course you need a bussard ramjet... to slow down.
    Here is a copy an article I wrote on a proposed "mission" - my hope is that we discover the Axion so that we can use that when "boosted" to 0.3c for relativistic flight.

    Combined Pulsed Deuterium Fusion And Axion Ramjets for Manned Deep Space Missions

  15. That begs the question, if we are using linear magnetic motors for generating power, I think we could also use them as suspension systems. If I remember correctly there was an automotive system shown here:,ride%2520performance%2520and%2520comfort%252C%2520vehicle&ved=2ahUKEwjVpMqnz_jvAhWSVs0KHQFkBZYQFjAOegQIBBAF&usg=AOvVaw3JXKshYOz-kMzZKYo7W3xr. So my thought perhaps, would be to use linear electromagnetic suspension and power generation in a combined mass saving system. The motor would push against the plate to provide suspension, and then the explosion would push the plate inward to generate electricity. This provides a very responsive highly controllable pusher plate that can be scaled for optimum performance. Just my two cents.

    1. Steel springs are robust, cheap and strong for their weight. Getting a magnetic system to the same standard of performance is going to be difficult... until technology progresses!

  16. Have you considered having the suspension arms double as linear alternators, with the kinetic energy the convert into electricity assisting the other shock absorbers?

  17. Do you even need pistons if its unmanned?
    Whats the max acceleration possible(reasonable speed you can deploy bombs) if you are willing to evaporate/damage the plate because its a Missile and going to destroy itself anyways?

    1. If you don't care about holding a squishy human crew inside the spacecraft, then a solid mount for a pusher plate can get the job done. There would be bonebreaking shocks with each nuclear blast, but it means that you can accelerate as hard as your structure can handle.

      For a historical example, the SPRINT missile accelerated at 100g. An Orion with that acceleration can get to 10 km/s in 1.7 minutes.

    2. Ten seconds, genius.

  18. Great article, as usual!

    I've been doing some heavy research in looking at the Orion drive for a generation ship for travel to Proxima Centauri, and I have a few concerns. Using Dyson's paper on interstellar travel, he notes that an ablative pusher plate would take around 133 years with a 1g acceleration (3.3% speed of light).

    The question is how do you continuously coat the pusher plate with oil or other substance while in space to prevent ablation with an explosion occurring every 3 seconds.

    The second question is how would you mount a rotating torus (for human passengers that is equipped with soil and water) without it breaking apart from the entire ship?

    1. Hi John!

      Doing a bit of maths, it seems like that generation ship accelerates for 11.7 years at 1g and then drifts the rest of the time.

      The Orion proposals that sprayed their pusher plates with oil did so every 0.6 to 1 second, so managing it every 3 seconds should be no problem. Think of it like the injector in a car piston spraying fuel into a cylinder. It reliably coats everything in oil within milliseconds - this is just the large scale version of that process.

      For the torus: consider a very large torus that has the Orion drive strung up on tethers within its center. It would basically pull along the torus while being suspended by the tethers, which makes the transfer of forces gradual and gentle.

  19. Thanks so much, MB! We ought to collaborate on a sci-fi book sometime.

  20. I have an harsh question, what will happen if not accelerating?
    The concept is quite solid during the use of the engine but the energy is a byproduct of thrust so never on-demand, seems like we must consider an enrgy storage element for a complete picture.
    If we store is possible to invest in some exotic mix like a second kind of explosion drive based on pure fusion that have negative energy balance, use fission-fusion to charge than discharge as pure fusion. For weapon i suppose is too difficult to get advantage from the storage and peak energy output is quite in line with the use because during combat acceleration matter still storage is a bonus for ancillary systems that are power hogs: railgun and laser get the most but you have to coope with RCT, radar, hotel and probably a PD, each one use a fair amount of power but in a less impulsive fashion.

    I suppose that our Orion ship will take a lot of time just drifting around or in parking orbit, so real ship will probably complement the energy of the drive with solar/nuclear and keep a fair amount of radiators anyway so devote some of this to keep a good reseve of liquid nitrogen is probably a good strategy not only for the drive but also for more common use, a good heatsink is always useful on a spaceship and open opportunityes for other cryogenic systems aboard.

    1. Hello!

      I addressed this in the blog post. There are two main uses for the Moto-Orion: military and ignition.

      A military spacecraft might need a lot of power for equipment such as radar or lasers. If it needs to use these, it also must move around to dodge projectiles or chase a target. A Moto-Orion is a perfect solution as it provides both thrust and electrical power at the same time.

      Some propulsion systems need a lot of electrical power to run. For example, the Mini-Mag Orion used huge electrical pulses to compress small pellets of nuclear fuel to ignite them. This is the second main use for Moto-Orion. It allows for power to be extracted for igniting nuclear fuel while also converting nuclear explosions into thrust.

      What you seem to be concerned about is situations where you need a lot of power but do not need to produce thrust. I would say that these are rare. And if they are unavoidable, using the Moto-Orion and then cancelling out the acceleration by thrusting in reverse later is a potential option. Orion-type spacecraft have plenty of deltaV to spare anyways!

    2. Sorry, i missed a not: "For weapon i suppose is NOT too difficult...." !
      Still the mass budget ask for some quite sizable storage, i suppose in the form of some fast charge and controlled discharge system like supercapacitators for get rid of the direct coupling between the huge pulsed output of the drive and the needs of other alements: the "power stroke" of the drive generate a lot in the first acceleration of the plate and a decreasing amount during the slowdown creating something like a sawtoot pulse of huge power and followed by a recover time during the pusher bounce.

      My intention was refining this bright idea about the Orion concept, with your regenerative braking i suppose is quite possible to put the Orion approach on pair with Medusa as a relatively lightweight approach that is usable today without exotic or fancy tech.
      A lot of saving you did is getting away the mass penality of various "springs" in the damper and making it quite light using EM capture, you got rid of a lot of operational issues decoupling the plate positioning from the rebounce and ideally opened the option for a one pulse use flexibility. Quite a lot of improvement!
      Bonus from your Orion is quite a compact (pusher retracted) and nimble (a lot of power for RCS) ship, comparing to Medusa the drive is quite clean and unobstructive to sensor and weapons and give a lot of options as dual use: chasing an Orion is living in his exaust, not many missiles can survive a nuclear blast and probably more you get close more push you get from the plasma of the explosion, the boom "push aside" spherically ceating a virtual wedge that must be compensated for.... no dumb fire in the back!
      As rule i didn't see an hard scifi future with a lot of spaceships combat, your option is in line with my view of a situation similar to sail navy with few ships of the line that can vaporize almost anything but too blunt for almost anything beside a total war and a lot of interesting action based on relatively small and unimpressive ships that can do boarding and policing actions: off planet destroy is quite simple but you are wrencing the assets you desire and usually it's not what you are looking for, juicy targets probably are not interested in heavy defence outside the basic needs for space hazards because their goal is also keep damage low and budget down, military targets probably are defended but most of the value is in data and tech that can be taken from the scraps or the proyection potential they give.

  21. A look at the first light fusion system ( ) might be interesting relative to Orion. They are developing a projectile impact triggered fusion power system where the fusion fuel pellet is imbedded in a larger clear plastic impactor target. Firing a small disk at the target at something in the 10-20 km/s rage sends shockwaves through the impactor plastic, into the fuel as a symmetrical implosion. So for a Orion drive, your droping a small cube, and firing the disk at it with a coil gun or something, which makes for a small light Orion.

    1. Yes! I've been following their progress for several years now.

  22. The concept is quite solid during the use of the engine but the energy is a byproduct of thrust so never on-demand, seems like we must consider an enrgy storage element for a complete picture.