Atomic Rockets

Sikon's Analysis

For a broad overview of some of the issues, study this penetrating analysis by the man known as Sikon

Space Warships: Power Generation, Waste Heat, & Firepower

Weapons like particle beams and lasers may have "unlimited ammo" if a space warship's electrical power generation and storage system is powered by nuclear reactors, with gigawatts or more of firepower.

Future ultracapacitors could have an energy density higher than 60 Wh/kg along with a power density greater than 100 kW/kg. Such is from a MIT study on ultracapacitors for future cars, implied here. That would be up to 0.2+ TJ of electrical energy stored per 1000 metric-tons of ultracapacitors, able to be discharged at a rate of 0.1+ TW. For example, a 100,000-ton warship with just 5% of its mass as ultracapacitor banks could store a terajoule, then discharge it at a rate of half a terawatt. Technology of the distant future may be superior, but the preceding is a reasonable lower limit. Energy storage is not the only limiting factor, though.

(ed note: Anthony Jackson thinks that 60 Wh/kg should be considered a high end estimate, not a low end. He further notes that 100x is approximately the theoretical limit for energy storage with chemical bonds, and as noted, 5 kilotons of capacitors hold 1 TJ.)

What is the recharge rate from warship power generation? The energy content of fission, fusion, or antimatter fuel can matter less for the attainable electricity generation than engineering limits. Even before melting, metals weaken if temperatures rise from more heat transfer into them than coolant systems take away; parts deform if subject to excessive mechanical stress; etc. For example, plutonium "fuel" in a bomb allows a power-to-mass ratio of billions of gigawatts of heat and radiation per kilogram during the fraction of a microsecond of detonation, but that of a plutonium-fueled power plant must be orders of magnitude less. A nuclear-electric concept with a MHD generator was estimated (PDF file) to obtain 0.37 kg/kWe, which would be 2.7 MW/metric-ton. For perspective, car engines of today are sometimes hundreds of kW of mechanical power per ton (i.e. 200 hp engine = 150 kW), with aircraft engines up to much higher power density. Even with need for electricity rather than mechanical power alone, the many thousands of tons involved in a space warship would allow it to have nuclear power generation at least in the gigawatt range or higher, likely terawatts for large ships. There would also be inefficiencies.

What about waste heat? Deploying large radiator panels while firing weapons wouldn't be desirable. Internal phase-change-material (PCM) heat sinks like ice/water might temporarily absorb heat. Actually, if the space warship has structure, armor, and individual weapons massing thousands of tons, such could absorb some gigajoules to terajoules. But such could not sustain a high rate of fire for long without needing a "cooling off" period, so a different system would be needed, at least as a supplement. The preferred radiator design for an armored warship is a droplet radiator, a charged (solid) particle radiator, or another alternative to large, vulnerable panels.

Radiator mass for the weapons is going to depend much upon acceptable operating temperature. If most parts of the weapons can operate at moderately high temperature, the waste heat from high power consumption can be transferred away fast enough without excessive radiator size. One study of what is obtainable for heat rejection (PDF file) in space with merely today's technology indicates that 30 MW of heat could be dealt with by a 45 metric-ton Curie point radiator ( CPR) or by a 29 metric-ton liquid droplet radiator, for an average temperature of 380 degrees Celsius or 650 K. The space warship would operate at least in the gigawatt range, with orders of magnitude greater heat rejection from its weapons, but it could afford to have orders of magnitude greater radiator system mass. And it would be more advanced, higher-performance technology.


Mr. Andrew Jackson disagrees with Mr. Sikon's analysis.

The 0.37 kg/kWe reactor described has a heat dump at 150K. That's not practical for a spaceship; you're not going to run your radiators at 150K, nor are you going to use liquid nitrogen as a heat sink. It also has an efficiency of 22%. Using a higher temperature heat dump will reduce efficiency or power density (or both); in practice the heat dump has to operate at the same temperature as the radiator. Assuming a 10 GW reactor, it's likely going to have a heat output of 20-40 GW.

Let's assume that we can get a reactor with a 1000K heat output and an efficiency of 20%, with a power density of 1 kg/kWe. A 10 GW reactor produces 40 GW of heat. A perfect blackbody at 1000K has a heat output of 56.7 kW/m2, so we need about 18 square meters per MW, or 1.4 million square meters. A perfect blackbody could radiate from both sides, but if we're using a non-solid radiator of real materials it's not a perfect blackbody, so we'll just have a wing with an area of 1 million square meters. Assuming our ship is 200M long, that means the radiator wing is 5 kilometers long.

I think not. So much for heat radiators. Let's shift over to heat sinks. We'll use water, since it's easy to work with. This gives us a heat sink at around 400K, so we'll double our efficiency; a 10 GW reactor now produces only 15 GW of heat. Without vaporization, cold water can hold about half a gigajoule per ton; 10,000 tons could hold 5 TJ (if we store a slurry of ice, increase by 50%). If we allow the steam to vent (which more or less requires dumping it to space; you need a phase change, which means you can't keep the water compressed) we get another 20 TJ.

Now, a laser system that would make the military jump for joy would have a peak output of 1 kW/kg, an efficiency of 20%, and a duty cycle of 20%, for a mean power output of 1 kW/kg. It will produce low temperature heat, well suited to our water sinks (and nearly impossible to radiate away with our high temperature radiators). A 10,000 ton weapon system requires a power input averaging 10 GW, and a peak power input of 50 GW. If we can fire for 15 seconds before triggering a cool down cycle, we need capacitors good for 40 GW * 15 seconds or 0.6 TJ, so 3,000 tons is all we really need.

During 1 duty cycle we produce 800 GJ of waste heat from the weapon. Generating 1 TJ produces another 1.5 TJ of waste heat, for a total of 2.3 TJ. We'll round up, and discover that we can run through 2 15-second duty cycles without venting coolant, and another 8 by venting coolant. Our combined system mass is 33,000 tons.

Now, if we have some down time, we probably want to bring the coolant temperature down to near freezing, or if possible turn it into an ice slurry. Unfortunately, that means a radiator operating at an average of about 300K, with a heat output of 0.46 kW/m2. If we figure extended radiators are 1 km long and 200m wide, they can dump heat at a rate of 180 MW, or approximately 8 hours to cool to near freezing. Generating the ice slurry would take another 6 hours or so. Also, unlike high temperature radiators, sunlight heating the radiators will interfere substantially with cooling, so we need to remain edge-on towards the sun.

Andrew Jackson

Back to Sikon:

Let's add an intuitive illustration of the overall picture. Consider 10% of the mass of a 100,000-ton warship being a beam weapon, with the maximum energy it could fire per shot or in a second being somewhere between 0.01 TJ and 1 TJ. That proportionally corresponds to as much firepower per unit mass as a half-kilogram energy pistol firing shots between 500 J and 50 kJ of energy. Such is equivalent to the energy pistol being able to vaporize a volume of ice between 0.7-cm and 3.3-cm in diameter per shot, like vaporizing a ball of ice between the size of a pea and a golf ball. While the whole range is conservative by sci-fi standards, one could take the low end of the range if concerned about the reliability of it being plausible. The comparison is proportional since the sample space warship's weapon masses 20,000,000 times more than the energy pistol.


Rick Robinson had an observation:

Big Proviso: (Sikon) is talking in terms of big ships; the example they give is a ship with mass of 100,000 tons, presumably "Washington Treaty" mass, not including remass (propellant/reaction mass). This is roughly the size of the largest ships I think are provided for in Attack Vector: Tactical. It is about 10x the mass, from my impression, of the largest type (DiGleria?) in regular service in the AV:T Ten Worlds setting.

By my rule of thumb such a ship would cost (the societal equivalent of) some $100 billion. (YMMV!) Mid-future colony worlds of the Ten Worlds or Human Sphere type, with populations no more than ~100 million, would be hard put to have more than a showboat or two of this class. (Sikon) speaks of fleets with thousands of such ships - so they're implicitly dealing with vast galactical-imperial scale polities.

I've gone into this a bit because it makes an interesting point: scale matters. I didn't carefully examine (Sikon's) analysis, but it gave the impression of being well thought out, and I can imagine that you could indeed get Incredible Firepower ... if you can afford an Incredibly Huge And Costly Ship.

Rick Robinson

Back to Sikon:

Yet the warship's shots each correspond to the equivalent of approximately between a 2500-kg high-explosive bomb and a 0.25-kiloton tactical nuke in the energy delivered. Beam weapons of such energy can have "unlimited ammunition," powered by the discharge of the capacitors, which are recharged by the warship's nuclear reactors to fire thousands of shots in a period of a few hours. Or smaller shots could be used for an even higher firing rate. For example, if a warship can fire a single concentrated 0.01 TJ to 1 TJ laser shot in a second, it might alternatively have weapons capable of sending out equal energy in the form of 100,000 to 10,000,000 one-hundred-kJ pulses per second over a huge shotgun-like pattern to hit a target at much greater range than would be likely otherwise. What was optimal could depend upon factors including the type of target, but the attainable firepower is vast.

For perspective, a 100-kJ vehicle-mounted laser concept is considered by the Department of Defense to be lethal against common rockets, aircraft, and light ground vehicles (PDF file) with little armor. Yet, at the technological level implied by sci-fi interplanetary or interstellar space war, average firepower of a far larger space warship could be astronomically higher, either in the energy per shot, the number of shots fired per minute, or a combination of both. Every 0.01-TW of average weapons power corresponds to 400 million times the energy per hour.

Propulsion system power could be much greater than electrical power and beam weapons power. For example, the MS Word document from researchers here describes a magnetic compression pulsed fission concept with a magnetic nozzle, in which a vehicle of 1310 metric tons initial mass and 100 tons final mass could have 263 GW jet power. That is between 0.2 GW/ton and 2.6 GW/ton, with relatively straightforward technology. For this distant-future scenario, such is just a probable lower limit. A much larger 100,000-ton space warship could be more than 1 GW/ton, corresponding to an exhaust jet power above 100 TW.

Beam Weapons: Planetary Assault & Space Combat

As an initial beam weapons illustration, consider a space warship firing a lethal radiation beam against planetary targets including aircraft. Against humans, on the order of 10 kJ per square meter of some types of radiation would be enough to cause enough exposure for relatively quick mortality, much above the level for slow death. The end result is a little like the effect of the radiation of a neutron bomb, for which 8000 rads or 0.08 kJ/kg-tissue (80 Grays) are enough to immediately incapacitate enemy soldiers like tank crewmen according to an U.S. military estimate, a couple orders of magnitude above the dosage usually lethal over a longer period of time (1% as many neutrons = 80 rads = 800-1600 rem in long-term). But the radiation wouldn't be neutrons.

This is not an ordinary particle-beam weapon concept, being instead a wide beam with particle composition and energies chosen to equal or exceed the atmospheric propagation of penetrating natural cosmic radiation. As GeV energies are obtained in contemporary research accelerators, the preceding would be attainable by an accelerator within a large space warship. Natural cosmic rays are 16 rem/yr in interplanetary space, dropping to 0.027 rem/yr at sea level. Since natural cosmic radiation experiences such an attenuation factor of 600 going through earth's atmosphere from space to ground at sea level, assume the wide-beam radiation should have an intensity on the order of 6 MJ/m2 before entering the atmosphere.

The result is that each shot of 0.01 TJ to 1 TJ energy can deliver a pulse of quickly lethal radiation to an area around 46 meters to 460 meters in diameter. If a given intensity level is insufficient, such as firing on a relatively hardened unmanned target, making the beam more narrow by a factor of 10 would increase the intensity by a factor of 100, and so on. But wide beams can kill ordinary tanks, aircraft, infantry, etc. The beam is unaffected by weather and sufficiently penetrates the mass shielding of the atmosphere, despite it being 10 metric tons per square meter. Unlike even neutron bombs, the beam would have no blast and just a few degrees heating effect when fired in wide beams, leaving structures unharmed aside from disruption to electronics, yet killing the occupants.


Again Mr. Jackson begs to differ:

Now, when talking about targeting the ground with a particle beam, it's worth noting that cosmic rays not only attenuate on hitting atmosphere, they scatter. You can't really target a region smaller than about 100 meters radius (31,000 m2). The attenuation length of cosmic rays at ground level is a bit over 100g/cm2, so 1 kJ/m2 produces a dose of about 1 gray; however, the radiation involved has a RBE of around 2, so it's about 2 Sv. Prompt incapacitation requires about 50 Sv (more vs rad-hard electronics, way way more vs bunkers), so we need at least 25 kJ/m2, or 800 MJ at ground-level, or 500 GJ at top of atmosphere. One duty cycle from our gun above is 150 GJ (15 Sv at ground level), and we probably don't want to dump coolant on secondary targets, so we likely only fire once or twice. In practice, the lethality difference between 15 Sv and 30 Sv is negligible (in either case, nausea after 5-30 minutes, a couple days of normal activity, then delirium and death), so one shot is fine.

It is also the type of thing that gets called a war crime.

Andrew Jackson

Back to Sikon:

Lethal radiation beams may also be used against other spaceships, with effectiveness determined in part by their shielding (armor) thickness. The extreme case is firing against a thin-hulled ship, in which case the attenuation factor of 600 for the previous scenario of firing through the 10,000 kg/m2 mass shielding of the planetary atmosphere doesn't apply. In that case, a quickly-lethal 0.01 TJ to 1 TJ shot can be up to about 1.1-km to 11-km in diameter. Actually, since enemy vessels can be detected at great range, the warship might not wait but rather open fire on lightly-armored targets at such extreme range that beams hit only by being hundreds of kilometers in diameter or more. The cumulative radiation dose delivered over many shots every minute would add up to enough in time. One potential countermeasure is mass-shielding or thick armor around vulnerable areas of a ship, like the battle stations for the crew and vulnerable electronics, such as with enough meters of metal to stop practically all of the radiation.

Another weapon can be microwaves. Against non-hardened civilian targets, as little as a few joules per square meter or less can be enough if delivered in the right time frame, concentrated into microseconds or less. Gigantic "EMP" pulsed microwave beams can fry ordinary electronics over up to many square kilometers per shot. EMP beams could be about the opposite of lethal radiation beams, devastating planetary infrastructure without killing any people aside from a few indirect deaths like crashing aircraft. Against more hardened targets, more focused microwaves in the form of narrow-beam MASERs might physically overheat and destroy. The potential firepower of such a concentrated MASER beam is implied by the many-gigawatt or terawatt-level power generation of a large space warship being equivalent to a number of tons of high-explosive per second.

As implied by what happens to sunlight, light from space doesn't always reach the ground well on cloudy days. Thus, lasers might be an unreliable weapon against planetary targets, unless the basic principle of this could be applied with ultra-intense pulses. However, the situation is different in space against enemy warships. The shorter wavelength of lasers compared to microwaves allows a more narrow focus at long range.

Projectiles & Missiles: Planetary Assault

During planetary attack, yet another potential weapons system for space warships is firing non-nuclear mass driver projectiles and missiles to hit air, sea, and ground targets on the planet below, impacting at hypersonic velocities. A 1977 NASA Ames study referenced here determined that an earth-launched mass driver projectile going up vertically could pass through earth's atmosphere from ground level to space with a few percent of its mass being an ablative carbon shield, losing only 3% of its total mass in the transit. Such is for a telephone-pole-shaped projectile of a metric ton mass. That means the reverse is also possible for projectiles with the right mass, dimensions, ablative shield, and trajectory. For example, consider a similar projectile fired from space, reaching the upper atmosphere at 12 km/s velocity and going nearly straight down. It could hit a ground target at about 11 km/s, a kinetic energy equivalent to about 15 tons of TNT explosive.

Projectiles and missiles fancier than the cheapest unguided shells could use small thrusters to adjust trajectory to home in on a target. Although sci-fi sensors or even remote-control communications systems might be able to operate through the plasma sheath from atmospheric passage (i.e. using high-frequency pulses of directed radiation or particles rather than ordinary radios), the simplest solution is if it instead slows down to a lesser Mach number first. Advanced robotic missiles tracking by the right combination of infrared, visible, radar, and/or other sensors could be hard for planetary targets to evade ... although space warships could alternatively just use their beam weapons against those targets.

Projectiles & Missiles: Nuclear Weapons in Planetary Assault

Large numbers of nukes may be used in planetary assault. For example, one cheap "brute force" method of dealing with atmospheric fighters trying to avoid shells or missiles might be to have them explode with sub-kiloton to single-kiloton yield. The equivalent isn't done by terrestrial militaries for reasons like political issues, but those do not necessarily apply so much in a sci-fi planetary assault scenario. Even in the real-world today, nukes do not have to cost more than merely hundreds of thousands of dollars each or less in mass-production, compared to fighters costing orders of magnitude more: tens to hundreds of millions of dollars each.

Fallout from such nukes would tend to be harmful to the planetary defenders and localized regions without making the planet unusable by the invaders. Localized radiation levels shortly after a detonation can be lethal, but such decrease over time. The radioisotopes emitting the most initial radiation are those with the largest fraction of their atoms decaying per unit time. (The rate of radiation emission per unit time from a radioisotope is inversely proportional to half-life, to a degree such that stable elements can be thought of simply as those with infinitely long half- lives). Compared to residual radiation one hour after the detonation, radiation levels are 1% as much after 2 days and 0.1% as much after 2 weeks. The fallout of a nuclear weapon detonation of low or moderate yield can much elevate radiation levels over a limited number of square kilometers, but it can do very little overall over the half-billion square kilometer total area of a planet like earth.

Historical above-ground nuclear weapon tests in the 20th century amounted to 440 megatons cumulatively, with 189 megatons fission yield ... 189000 kilotons (large PDF file). Total collective dosage to the world's population from such past tests corresponds to 7E6 man-Sv, for the UNSCEAR estimate for total exposure in the past plus the result of currently remaining radioisotopes projected up through the year 2200 (PDF file). The preceding total over the decades and centuries is less than what is received every year from natural sources of radiation (PDF file), which is in turn orders of magnitude less than what would make an eventual death from cancer probable. Of course, from a real-world civilian perspective, any potential increased risk of cancer is undesirable, but, from the perspective of the hypothetical space invaders, the bulk of the planetary surface is not harmed enough for them to necessarily be concerned.

For example, even with fission devices, if the orbiting warships are firing quarter-kiloton-yield nuclear shells or missiles against targets like enemy aircraft, it would take on the order of 800,000 warheads even just to exceed the limited radiological contamination from the 189-MT fission component of the preceding nuclear tests. If available, pure-fusion devices would be cleaner. Sci-fi technology allows other possible ordnance, such as biological weapons genetically engineered to have a non-lethal temporary incapacitating effect or infectious nanobots. Different attackers might use different techniques depending upon their psychology, ethics, objectives, etc.

Missiles vs. Point Defenses in Space Combat

In combat between space warships, the vast firepower attainable from nuclear projectiles or missiles, combined with no particular limit on range, might make them dominate the battlefield. Or they might not, depending upon the effectiveness of missiles versus point defenses, their relative cost, and other factors in a given sci-fi scenario. With lasers destroying artillery shells becoming possible even now, the point defenses of distant-future space warships are not to be underestimated.

As little as a 100-kJ projectile can destroy an ordinary missile. (For perspective, 100-kJ is like the kinetic energy of a 200-gram projectile going 1 km/s, although the analogy should not be taken too far since the momentum is different for a much higher velocity but far smaller projectile). For example, if warship firepower of 0.01 TJ to 1 TJ per second is attainable as previously suggested, such could allow a mass driver or mass driver array firing a 0.01-GJ to 1-GJ shot per millisecond. If firing pellets like a shotgun, such could deliver on average a 100-kJ pellet per square meter within a 11-meter to 110-meter diameter pattern per millisecond, a thousand times as much per second, potentially destroying many different incoming missiles. Or, to maximize engagement range, firing a whole second at one target could amount to a shotgun pattern 0.36-km to 3.6-km in diameter. Alternatively, comparable firepower to the preceding might also be obtained with another weapons system like a laser array instead.

Against such point defense firepower, ordinary missiles are at a disadvantage against warships. Still, if the missiles aren't so ordinary, there may be countermeasures to point defenses, such as faster, more armored, and/or more numerous missile swarms. One possibility could be a space missile swarm not carrying sizable nuclear warheads but rather dispersing clouds of kinetic-kill masses, such as billions of grains of sand or the equivalent, too numerous for point defense weapons to hit and vaporize them all. Point defenses might try to destroy such missiles far enough away for clouds deployed before missile destruction to subsequently miss due to the warship's changing course.

The Value of Weapons Range, Mobility, Armor, & Point Defenses

Imagine two modern-day soldiers. One is armed with a sniper rifle, while the other is armed with a pistol. If they face each other in a jungle or in dense fog with visibility not beyond several meters, either one may have a good chance of being the winner. But now imagine them starting a kilometer apart on a featureless flat plane of solid rock with perfect visibility. Then the guy with the sniper rifle wins, as the man with the pistol can not approach close enough to hit before being shot by the sniper. Since there is typically no effective stealth in space, the situation for warship combat can be like perfect visibility, no horizon, and usually no cover. That makes effective weapons range particularly important.

Fire control computers try to predict a target's position based on its velocity and current acceleration, but, at ranges with significant light speed lag, mobility matters much against beam weapons (and possibly the missile-deployed kinetic-kill clouds described earlier). For example, a ship doing 5g of unpredictable acceleration deviates 25-m in 1 sec, 2.5-km in 10 sec, 88-km in 1 minute, and so on. One countermeasure may be to fire many shots, but the earlier illustration of a warship firing a huge pattern of 100,000 to 10,000,000 100-kJ shots per second doesn't work well if the target has armor making 100-kJ too little. Armor could make the enemy fire a low rate of concentrated high-energy shots, reducing the chance of any hitting at long range. Of course, good enough point defenses are also needed, or else the armor would just be penetrated by a missile with a nuclear warhead.

Defeating Anti-Space Weapons

What about space warships fighting planetary anti-space weapons? Typically the planet would be better off having space warships than planet-based weapons. Launch a missile from a planet with a regular rocket, and more than 90% of its mass is involved just getting off the planet. Even if an advanced propulsion concept like nuke-pulse or nuke-saltwater rockets is used instead, having such launched from a planet during a battle would make them relatively easy targets during boost phase.

Craft launched from a planet may tend to be smaller and more limited than space warships. For example, a mass driver sending even just ten tons per hour to orbit could over a decade put almost a million tons up, enough to be potentially the seed of a society processing eventually billions of tons of extraterrestrial material into habitats and ships. But, in that scenario, billions of tons of spaceships might exist without the planet necessarily being able to launch more than a proportionally minuscule amount in a day. There is likely shipment off-planet of some valuable goods and also passenger traffic, but X million people per decade going off-planet only corresponds to just 20 * X * Y tons per day needed, where Y is the ratio of total launch mass to body mass.

A planet could have gigawatt to terawatt range beam weapons, but the effective range of such against space warships would tend to be less than vice versa: In a duel at up to light-minutes or greater range with light speed weapons, a space warship fleet will tend to win against a planet, as the immobile planet with zero unpredictable acceleration can be engaged at extreme range.

For example, if technology allows a variant of the lethal radiation beam weapon described earlier to have 0.1 to 10 microradians divergence, the beam would diverge 0.01-m to 1-m per 100,000-km distance, hitting a spot 100-m wide at 10 million kilometers to 1 billion kilometers range. With thousands of 0.01-TJ to 1-TJ shots fired per hour with electricity from the nuclear reactors, enough hitting a planetary target sooner or later, warships could devastate appropriate parts of the planetary surface from up to light-minutes to light-hours of range. That gives the mobile warships plenty of time to evade any light speed weapons fire from the planet. Such would arrive long after each warship has moved to another location in the vastness of space, perhaps millions of kilometers away from its previous position.

If even more firepower is needed, kinetic-kill clouds might be used, i.e. billions of particles of debris that defenses could not stop. For example, ships with nuke-pulse engines able to carry and send "cargo" on the right trajectory at 100 km/s to 1000+ km/s velocity could indirectly deliver 1,200 to 120,000+ megatons of destruction per million metric tons of material carried. Optionally, the columns of fire in the atmosphere created by the preceding might "blind" remaining defenses for critical seconds while missiles with nuclear warheads arrived right behind them. Before inefficiencies and aside from the other mass in nuclear weapons, fissioning plutonium and fusioning lithium-6 deuteride are 17 million megatons and 64 million megatons respectively per million metric tons mass. Of course, if the goal is to capture the planet with it still inhabitable, the level of firepower used in destroying anti-space weapons from extreme range would need to be limited. Warships could afterwards move closer, into orbit, providing final fire support for an invasion.


Isaac Kuo questions some assumptions:

I was struck by how it assumed the space ship had amazing beam weapons capable of penetrating the atmosphere, but for some unknown reason ground defenders using that same beam weapon technology simply lose.

On, the contest between beam weapons on mobile warships vs beam weapons on planets is completely lopsided in favor of the planetary defenders. They have a stupendous advantage in heat rejection, shielding, and mobility. Sikon ... ignores the heat rejection advantage and ... assumes the planetary systems can't use any shielding other than the atmosphere. He seems to assume planetary defenses must be fixed, despite the explicit example of aircraft which can literally jink all week (which, of course, spacecraft can't). Never mind about submarines, ships, or underground weaponry.

Isaac Kuo

As does Rick Robinson:

And the most important one of all, IMHO (though you may be subsuming it under mobility): stealth/concealment. A habitable-planet surface is about as cluttered an environment as you can find. Other parts of the post also seemed to blow off the problem of detecting targets on a planetary surface.

As an aside, at least "guns" reveal themselves when they fire. Assuming you have a suitable tech for lobbing missiles out of a gravity well, a missile engagement is even more in favor of the surface, because once a missile is fired all it leaves behind is its launcher, probably of insignificant value as a target.

Returning to beams, the whole sensor-blinding issue also heavily favors the planet, because finding a passive sensor on a planet surface approaches the level of trying to find a guy with binoculars somewhere on the nearside of the planet.

Rick Robinson

Back to Sikon:

Planetary Assault: Close Fire Support & Utilizing Recon Drones

With good enough targeting information transmitted from recon drones through a computerized system, space warships could help kill even individual vehicles or even individual enemy soldiers from orbit when possible. Such would not be their primary mission, and initially the warships would attack more valuable targets. But afterwards, a warship would still have practically unlimited ammo for its electrically-powered beam weapons running off nuclear reactors. Using a hundred-thousand-ton warship to kill a couple enemy soldiers riding around in a truck might superficially seem wasteful, but there is next to no marginal cost in the preceding scenario.

Consider a warship orbiting at 200-km low-orbit altitude for final fire support. A little like a terrestrial sniper can shoot an enemy from 0.5-km away, some beam weapons on the warship could be designed to hit precise locations on the ground below, with potential accuracy of within a meter. If there was a single person or handful of people on the warship manually trying to search for targets, aim, and fire the weapons, it would be a slow process. Yet, if there are a large number of robotic recon drones searching for enemy vehicles and soldiers, transmitting their precise coordinates, a computerized fire control system on the warship could shoot thousands of designated targets per hour, continuing for hours or days if necessary. Given the firepower and capabilities possible with one space warship, imagine what a fleet of thousands of such warships (or more) could do against a planet.

Space warships would initially destroy all targets they could see from space, but, for foreseeable technology, orbital surveillance might not find every last target. Deploying air and ground versions of robotic recon drones could help give further targeting information. For example, if a golf ball-sized robotic drone with a miniature jet engine flies up to the window of a building and sees enemy soldiers inside, it can transmit a signal causing the warship's computers to fry the area within a 50-meter radius with a lethal radiation beam a fraction of a second later ... potentially very effective yet still with less collateral damage than just nuking the whole city.

The preceding could be done before sending in regular armies or occupation forces in order to drastically reduce ground combat casualties, although use of expendable robots and/or telepresence whenever possible might make human or sapient casualties beyond non-sapient robots be low anyway.

The Unpredictability of Future Technology

Even in a hard sci-fi scenario, predicting the capabilities of technology that may be centuries or millennia beyond the 21st-century is highly uncertain. For example, perhaps technology would allow a million tons of raw materials to be quickly and cheaply converted to its mass-equivalence: a billion one-kilogram missiles to be dispersed at low altitude. Or there could be other weird military technologies. A little like a person from centuries ago couldn't very well predict the capabilities of modern combat, the preceding is mainly just a lower limit on what could be accomplished at the technological level commonly implied by interplanetary and interstellar wars in science fiction.


Ruppe's Analysis

Adam D. Ruppe had this analysis. It was in a thread at the Stardestroyer BBS.

I don't think there would be a huge variation in the types of warships seen. You'd have the big battleship which would dominate everything it fights, and then maybe smaller ships that could cover more area at once and engage in light combat, but wouldn't stand up to the battleships. Red called these 'frigates' in his Humanist Inheritance fiction, probably because their role is similar to the ship of the same name from the age of sail, and it is a term I like, so I will use it here. However, note 'cruiser' may also be an applicable moniker for these ships, probably depending on its specific mission rather than its design goal.

I feel these would exist due to economic efficiency rather than speed or range difference like those seen in the real sailing frigates. Let me explain.

Many of the arguments against space fighters can actually be used when talking about other capital ship classes as well. Let's look at what the roles of various naval ship classes basically were, and see if they could have an analog in space.

You had corvettes, which were small, maneuverable ships used close to shore. This role doesn't really apply in space. You might argue low orbit around a planet could be seen as a shore, but the problem is combat ranges would be rather large. If you have a stationary asset in LEO that you want to attack, you could put your battleship arbitrarily far away and attack it at will. If you have a mobile asset in LEO you want to attack, you can still attack it from some distance away, probably around one light second, to avoid too much light speed lag targeting issues and diffraction of your laser beams over the distance.

For comparison, the moon is about one and a half light seconds away from Earth. So, the battleship could be sitting out two thirds the distance to the moon and easily engaging the LEO target with precision and power. Corvettes being there wouldn't be of any help on defense, and the battleship can do their job on offense just as well, and at longer range.

A corvette type ship might be useful to the Coast Guard for police and search and rescue work, but that is an entirely different realm than a warship.

How about cruisers / frigates? The historical usage of the term referred to a small but fast warship, capable of operating on their own, and often assigned to light targets or escort duty. I do see an analog to this role in space.

A frigate would be no match for a battleship, however they would be useful in force projection, due to presumably being cheaper to produce and operate, thus more numerous. I'll be back to this in a moment.

And of course, battleships would be the backbone of the war fleet, able to swat down anything that comes at them except other battleships. If it were economically feasible to build a huge fleet of battleships, I see no reason not to. Let's investigate some of their traditional disadvantages and see if they apply in space.

The big one is speed: the huge battleship can take just about anything dished out to it and dish out enough to destroy nearly any other class of ship, but its huge size makes it slow. This isn't so much of a concern in space. Allow me to elaborate.

There are two things in space that are relevant when talking about "speed": delta-v and acceleration.

Delta-v is determined by the specific impulse (fuel efficiency) of the ship's engines and the percentage of the ship's mass that is fuel. Tonnage of the ship doesn't really matter here: it is a ratio thing. If the specific impulse is the same and the fuel percentage to total mass the same, any size ship will eventually reach the same final speed. Thus, here, if fuel costs are ignored, small ships have no advantage over large ships. (And indeed, if you are going on a long trip, the large ship offers other advantages in how many supplies or for war, how many weapons it can carry at no cost to delta-v, again, if the ratio remains constant) So the question is how fast can they reach it, which brings me to acceleration.

Acceleration is determined by total engine thrust and the total mass of the ship. At first glance, it seems that the smaller ship would obviously have the advantage here, but there are other factors that need be observed.

One is the structural strength of the materials of which the ship is constructed. This becomes a big problem on insanely huge ships with larger accelerations, since the 'weight' the spaceframe must support goes up faster (it cubes) than the amount of weight it can handle (it squares). Mike talks about this on the main site when he debunks the silliness of giant insects. However, steel is strong enough that with realistic sizes and accelerations, this should not be an issue before one of the other ones are.

One that is a much bigger problem is how much the human crew can handle. In the space / atmospheric fighter thread we had the week before last, Broomstick discussed the limits of the human body to great accelerations. Well trained people in g-suits can handle 9 g's for a short time, but much more than this is a bad thing to just about everyone - their aorta can't handle it. In fact 5 positive g's are enough to cause most people to pass out, as she explains. If the crew is passing out, the ship is in trouble. This problem can be lessened by the use of acceleration couches: someone laying down flat can handle it much better for longer, but even 5 g's laying down is going to be very uncomfortable, and the crew will have a hard time moving their arms. Extended trips would probably be best done at 1 g so the rocket's acceleration simulates Earth normal gravity, with peak acceleration being no more than 3-5 g's for humans in the afore mentioned couches if possible.

That is probably the most significant limit on acceleration, since it is an upper limit of humans. No matter what technology exists, this cannot be avoided.

The third limitation will be based on the technical problem of generating this much thrust for the mass. This, too, can provide an upper limit, since adding more engines on to a ship will eventually give diminishing returns. The reason for that is the available surface area on the back of the ship where the engine must go increases more slowly than the mass of the ship as it grows. But, for a reasonably sized ship, this should not be a tremendous problem, especially when nuclear propulsion techniques are used, many of which have already been designed and proven feasible in the real world. Fission nuke pulse propulsion can provide 400 mega-newtons of thrust according to the table on Nyrath's Atomic Rockets website (see the row for Project Orion).

Three gees is about 30 metres per second squared acceleration. F = ma, so let's see what mass is possible. 4e8 / 3e1 = 1e7 kg, or about 10000 metric tonnes. Incidentally, this is the number Sikon used for his demonstrations in the October thread about brick vs needle. I think it a reasonable number for a battleship, so rather than repeat the benefits of this, I refer you back to that thread and the posts of GrandMasterTerwynn and Sikon on the first page, who discussed it in more depth than I am capable of. I agree with most of the views Sikon expressed in that thread.

So, for these sizes, the speed argument against battleships is very much sidelined.

You also pointed this out later in your post that these advanced propulsion techniques do not necessarily scale down very well, which may also serve as a lower limit on ship size, which is probably more relevant than the upper limit it causes.

You might ask if pushing for a greater peak acceleration would be worth it, and it is not, in my opinion. The reason again goes to the human limitations. Even if your warship is pulling 10 gees, it most likely won't help against a missile, which can still outperform you.

An acceleration of even 1 g should be enough to throw off enemy targeting at ranges of about one light second. By the time the enemy sees what you are doing, you have already applied 10 m/s change to your velocity. Then, if he fires back with a laser, you have another second to apply more change. This would be enough to help prevent direct, concentrated hits. Having even five times more acceleration will offer little advantage over this in throwing off targeting or wide spread impact of lasers of particle beams, due to the ranges and the size of your warship, which is certain to measure longer than 50 metres. For missiles and coilgun projectiles, it matters even less, simply due to the time the enemy fire arrives, you have plenty of time - minutes - to have moved. 1g is plenty for that, attainable by a nuke pulse engine for sizes around 30,000 metric tonnes.

Long range acceleration would again be limited to around 1 g or less due to the humans, mentioned above. However, even at 1g constant acceleration (which would probably not be used due to fuel concerns anyway), an Earth to Mars trip could be measured in mere days. More offers little advantage there either.

Lastly, there may be a question of rotation. A more massive and longer ship would have a greater moment of angular inertia than a smaller ship, thus requiring more torque to change its rate of rotation. Again, I don't feel this will be a major concern. At the ranges involved, you again have some time to change direction. However, this does pose the problem in quick, random accelerations to throw off enemy targeting.

Going with the 10,000 metric ton ship, let's assume it has an average density equal to that of water: one tonne per cubic meter. For the shape, I am going to assume a cylinder, about 10 meters in diameter (about the same as the Saturn V), with all the mass gathered at points at the end. The reason of this is to demonstrate a possible upper number for difficulty of rotation (moment of inertia), not to actually propose this is what it would look like. Actually determining an optimal realistic shape for such a ship would take much more thought.

With this, we can determine the length of the cylinder to be 10000 / (π r2) = about 130 metres long. Now, we can estimate the moment of inertia, for which, we will assume there are two point masses of 5000 tons, each 65 meters away from the center. So moment of inertia for the turning axis (as opposed to rotating), is 2*5000 * 65^2 = about 4e10 kilogram meters squared.

Now, let's assume there are maneuvering jets on each end that would fire on opposite sides to rotate the ship. Let's further assume these have thrust about equal to that found on the space shuttle, simply because it is a realistic number that I can find: about 30 kilo-newtons. Let's determine torque, which is radius times force, so 3e4 * 65 * 2 (two thrusters) = about 4e6 newton meters. Outstanding, now we can determine angular acceleration possible.

Angular acceleration = It, where I is moment of inertia and t is torque. So, we have 4e6 / 4e10 = 1e-4 radians per second squared. This is about a meager 10th of a degree per square second. Remember this is acceleration - change in rotation rate. Once spinning, it would tend to continue spinning. This is also a lower limit: most likely, the thrusters would be more numerous than I assumed, and probably more powerful as well, and the mass probably would be more evenly distributed. But anyway, let's see if it might be good enough.

As I said when discussing linear acceleration, you would want some quick randomness to help prevent a concentrated laser beam from focusing on you, and you would want the ability to change your path within a scale of minutes to prevent long range coilgun shells from impacting. There isn't much you can do about missiles except point defense: a ship cannot hope to outmaneuver them due to limitations of the crew, if nothing else.

Some unpredictable linear acceleration should be enough to do these tasks, unless the enemy can get lined up with you, in which case, you will want to change direction to prevent him from using your own acceleration against you, and blasting you head on. So the concern is can you rotate fast enough to prevent the enemy from lining up with you. So, let's assume the enemy can change direction infinitely fast, and can thrust at 3 g's. The range will still be one light-second.

We can calculate how much of an angle he can cut into the circle per second if he attempted to circle around you. His thrust must provide the centripetal acceleration, so we can use that as our starting point. Centripetal acceleration is equal to radius times angular velocity squared, thus, sqrt(30 / 3e8) = 3e-4 radians per second.

So, its angular velocity is three times that of the acceleration of the battleship. Thus, it would take the battleship three seconds to match that rotation rate. It would also want to spin faster to make up for lost time, thus lining up on your terms again. I feel this is negligible because of two factors: if the enemy actually was orbiting like this, its position at any time would be predicable, thus vulnerable, and the battleship can probably see this coming: the enemy's tangential velocity must also be correct to do such a burn - he can not randomly change the orientation of his orbit due to his limitations on linear acceleration. This means you can see what he is doing and prepare for it with a small amount of time of him setting the terms. In this small time, he would not even move a degree on you: still easily within your armor and firing arc. (Also, weapons turrets on the battleship would surely be able to rotate at a much, much faster rate, so outrunning them is impossible anyway).

Thus, I feel neither linear acceleration nor angular acceleration are significant limiting factors as size increases within this order of magnitude.

Long story short: unlike marine navies, speed is not a significant factor in space warship design, unless you are getting into obscene sizes.

And, since I find it interesting, I want to finish talking about possible ship classes, so back to the comparison list.

Submarines depend on stealth, and since there is no stealth in space (barring pure magic like the Romulan cloaking device), there are no submarines in space.

Destroyers operated to protect larger ships against submarines and small, fast ships, like torpedo boats. Since speed is not a significant factor and stealth impossible, there are no fast ships nor subs, meaning the destroyer has nothing to do, thus would not exist. (Though, you might chose to call what I call frigates destroyers if you prefer the name, but IMO the role is different enough that is isn't really accurate. But the US Navy somewhat does this, so it is up to you as the author.)

A cruiser is simply a ship that can operate on its own. Frigates, destroyers, and battleships can all also be called cruisers depending on their mission.

A battlecruiser is a ship meant to be able to outrun anything it can't outgun - it had the speed of a lighter cruiser with the guns of a battleship. In real navies, this was usually achieved by taking armor off a battleship. However, since speed is not limited by mass in the given order of magnitude, a battleship and battlecruiser would have the same speed: the battleship would be a clearly superior vessel. Thus, no battlecruisers. (Now, if you have FTL, then that might create a battlecruiser class, but I am trying to avoid talking about magic in this discussion, since as the author, it is entirely up to you what the magic can and cannot do.)

A destroyer escort is a small, relatively slow ship used to escort merchant ships and protect them against submarines and aircraft. But, in the real world, aircraft can threaten a ship due to its superior speed and submarines due to stealth. So neither of them are there, making the destroyer escort worthless. Frigates or battleships would have to be doing the escorting, since they are the only things that can stand up to what they will be fighting: other frigates or battleships.

Now, a little more on what I mean by frigate. It is basically a smaller battleship, built simply because I am presuming they will be cheaper to produce and maintain, thus allowing more of them to exist. With more of them, they can be in more places doing more things. Cost is the only real benefit I can think of: if for some reason you could crank out and operate / maintain battleships for the same cost, I see no reason why you would not.

The 10,000 ton proposal might actually be the frigate, with the battleship being larger than that, or it might be the battleship with the frigate being smaller than that. The relationship would remain the same, however.

Adam D. Ruppe

Origin and Evolution of Space Warships

Arthur Majoor had this cogent analysis of the origin of space warships.

... One thing which I notice you haven't touched on is the origins and early development of a space navy. No one is going to be able to operate a heavy space cruiser the size of an OSCAR class submarine (much less the Polaris) without climbing a fairly steep learning curve.

Military access to space is probably the route that will take us there. The first step is already here; the recent introduction of operational ABMs. Assembly line production of ABM systems (since they will eventually be needed to cover both the east and west coasts of the United States, as well as the polar routes and deployed to protect bases in Guam, Diego Garcia etc.) should lead to standard "busses" and launchers for critical space hardware and will certainly drive down the price of getting into space and operating in LEO. The experience gained by assembly line production will increase the reliability of this hardware and associated systems.

The ever growing amount of critical space infrastructure and hardware will demand the ability to "surge" large numbers of satellites into space in response to a crisis or to replace damaged and destroyed assets in the early stages of a war, and one or more manned "garages" to service orbital hardware and extend its useful life. The ABM launcher will be produced in such quantities to make it the cheapest and most reliable vehicle for lightweight orbital hardware and many military and commercial systems will be designed to take advantage of this. Mass production of ABMs and their interfaces might make this the common standard for most space hardware into the future. Manned launchers and spacecraft to operate in LEO that are derived from these systems will share the cost and reliability attributes of the base system.

Given the base launcher is a solid fuel missile with a fairly narrow diameter, the eventual manned spacecraft will resemble the one man space cruiser concept from the 1980 era "High Frontier" proposals (i.e. a very minimal one man spacecraft), rather than the luxury yachts or tourist ships some people think would lead to space access. Since the one man "space cruiser" would have limited supplies and on orbit time from direct launch, military space stations would be required for in orbit refueling and replenishing. The space stations themselves will be pretty Spartan, given each section has to be sized to fit on the standard launcher in folded or deflated form. They may even operate unmanned for a large part of their life times to extend their limited supplies. Since these garages would be vulnerable to enemy ASAT weapons if left in a fixed orbit, they would have to be spacecraft in their own right, capable of manoeuvre and orbital changes, and at the very least treated to minimize visual, radar and thermal emissions.

While small military garages will be the starting point, eventually there will be a need and desire for larger and more capable systems. Clustering garage segments together to make larger space stations will be a first step. Several cylinders ganged together provide a sheltered "dock" for spacecraft, while the upper surface can be utilized for tank farms, solar panels and other systems.

For protection against space debris, inflatable wake shields will be common equipment on long duration space hardware. Using high density foam to "blow up" the wake shield and supporting struts, the wake shield is either made out of metalized Mylar film (civilian spacecraft) or "dark" materials which are absorbent over a broad range of wavelengths in order to render the spacecraft less visible. The adoption of inflatable Mylar wake shields provides the experience needed to create inflatable high gain antennas and optical mirrors. The mirrors can be used as concentrators for solar furnaces, solar thermal engines, solar power generators (either using photovoltaic cells at the focus or a thermal generator) and they can also be used as "one shot" fighting mirrors for ground and space based laser weapons.

As the space force grows in size and importance, the need to ferry larger amounts of supplies and create more capable space forces drives the development of the ARES heavy lift launcher, a "C-130" for space. Resembling an overgrown Space Shuttle External tank with an engine cluster, the ARES has hard points along the side to house payload pallets or strap on boosters, as well as an upper collar for nose mounted upper stages. The ARES launchers would be retained in orbit to become the building blocks for second generation space garages, larger space structures or as the elements of commercial or Space Navy spaceships. In time, these space stations/space ships might also be home to space bombardment weapons and a form of Glider Infantry using a variation of the SUSTAIN concept to insert forces on the ground in times of crisis.

The internal tankage of the ARES can be subdivided by inserting a series of balloons after orbital insertion and hardening them with epoxy. For space station building blocks this is sufficient, the ARES shell has hard points for payload attachment that double as connectors for attaching these units together. These attachment points also serve to connect various systems to the budding space structure, such as trusses, solar panels or the inflatable wake shield. Larger structures can be joined "nose to nose" and rotated about a common centre of axis to provide artificial gravity for the crew on extended missions. A permanent space station would have a vast wake shield in the direction of orbit, a series of ARES stages spinning on a common axis and one or more ARES units "trailing" for the docking station and zero g workshops, etc.

To create a spaceship, one or more units are connected, and a nuclear thermal rocket (NTR) is attached. Some of the tankage is retained to hold the reaction mass for the NTR; for logistical reasons the preferred reaction mass would be LOX from lunar sources. To gain access to this resource, ARES stages would be refitted and refueled in orbit and sent to the moon, where they become part of the ground infrastructure or serve as tankers to bring LOX back to Earth orbit. A crew cabin and landing legs are the normal retrofits for ARES lunar systems. While not as efficient as liquid Hydrogen, LOX is easily available from regolith, can be boosted at low expense from the moon, can be stored for long periods of time in orbit (since it is a "soft cryogen"), and also doubles as an emergency source of breathing oxygen for the crew. At least one of the balloons near the center of the ARES is covered in radiation shielding to act as the "storm cellar" during solar storms and other radiation events. The NTR must be made of materials which can withstand the effects of exposure to high temperature oxygen, or be a "nuclear light bulb" (since I suspect no one will allow open cycle nuclear reactors to operate in Earth Orbit or even Cis Lunar space).

As time goes on, Lunar resources become inadequate (since they are deficient in most light elements) and the moon itself would be considered vulnerable to attack from Earth. Greater strategic depth is achieved by establishing bases among the Near Earth Asteroids, since they provide both the needed light elements to sustain life and industry in space, and also provide time and distance to protect deployed elements of the Space Navy from rival forces on Earth and in space. This also allows the reaction mass to become water mined from asteroidal sources, since it is even easier to deal with than LOX. The simplified logistics suggests that "steam rockets" using nuclear light bulbs as the energy source and water as the reaction mass would be the system of choice for military and commercial spacecraft. Operations in deep space will create a different paradigm for operations (as so many people have noted). My take on the matter revolves around the requirement for protection from cosmic radiation and debris in transit, as well as the provision of artificial gravity for the crew(s). The cycler concept of having a large space station traveling between the planets is a good starting point, and the "Ice spaceship" provides many of the features a Space Navy might desire. The 215m diameter ice ship is protected by 40,000 tonnes of water ice, a vast thermal sink against energy weapons and a pretty hefty shield against kinetic energy impactors as well. Since the basic structure can be created with a 60 tonne "bladder" and filled with water extracted from asteroids, large numbers of cyclers can be assembled in space.

The actual warships would be clustered in the middle, connected to the structure to provide thrust and electrical power with their engines, while drawing water as reaction mass from the ice spaceship, or using the ice as a thermal sink for the engines and energy weapons while clustered together. The large size of the ice ship also serves as a means of supporting a large sensor array, which can support operations both while clustered with the warships, or when the mini fleet is dispersed. This sort of arrangement would work well when dealing with planetary sized targets like Mars or the Jovian moons, you have a base to maintain and preserve your ships to and from the theater, but can disperse independent warships with full tanks and weapons load when you reach the area of operations. (presumably the Fleet HQ has sent several empty ice ships to intersect the planet at different times so you can refit, rearm and go home, but the ships are capable of doing so independently if necessary). In effect, the Space Navy would be based on a series of "submarine pens" moving between planets.

This isn't such a far fetched analogy. I suspect the actual warships by that time would resemble an OSCAR class submarine in size and function. Even construction of the spaceship would be broadly similar, with the outer casing being used to house reaction mass (in the form of water) and the missile racks or beam weapon emitters, while the inner hull would probably be very small and heavily automated like an ALPHA class submarine. Since the crew would be inside the 214m ice ship during the cruise portion of the flight, they would have their gravity and cabins there, while in combat they would be in the warship cycling between zero g and "forward is up" orientation, spending their time strapped into acceleration couches.

For small targets like an asteroid, the smaller 100m ice ship might actually be the warship. It has 8000 tonnes of ice to act as a heat sink or reaction mass, is powered by three NTR's (according to the author. This could be reduced if more powerful nuclear light bulbs were substituted), and has lots of interior room for a large crew, supplies and so on. Rather than carry independent warships in the center cavity, it might be loaded with hundreds of missile busses; a space going arsenal ship. For mass fleet actions (i.e., conquering the Uranus system to seize the Helium 3 facilities), a combination of 100m arsenal ships and 214m cyclers carrying independent warships seems to provide the balance of firepower and flexibility a Space Navy would want.

Arthur Majoor

Mr. Majoor's proposed future history is logical and self-consistent. However, as with all analysis of this type, it does rely upon a couple of assumptions. People who want to alter the history can tweak the assumptions.

An interesting timeline, radically different from what I usually assume. I tend to assume civilian development of space will occur before development of military spacecraft beyond spy satellites.

Arthur Majoor's analysis starts with a couple ideas which would never even cross my mind. First, he assumes ABM missiles would use rockets suitable for orbital launch. I always assumed such missiles would lack the delta-v for orbit, but honestly I never checked.

Second, he assumes a future need to be able to "surge" orbital military hardware at times. I actually think the opposite to happen. I believe that improved technology would make each spy satellite more capable, reducing the number needed and thus reducing military launch demand. At the same time, I expect UAVs to become cheaper and more capable. Since they can fly underneath clouds, they can provide better tactical intelligence and better 24/7 coverage.

So, I'd expect a small number of spy satellites used for long term missions during peacetime, along with mostly aerial UAVs for a "surge" of intelligence capabilities during wartime.

Isaac Kuo

If you presume normal ABM tech in use, you also have to have a corresponding capability to surge useful satellites to replace those shot down by the other side. They kind of go together.

Mark Graves

ABM technology is different from ASAT technology. There might be some overlap in that a low Earth orbit satellite may be engaged with ABM missiles, but generally the tasks are very different.

An ABM missile doesn't need orbital velocity, but it does need to intercept its target on the first pass and it needs to intercept with a very high relative velocity. In contrast, an ASAT missile probably needs orbital velocity, but it can use a closely matching orbit to intercept at relatively low velocities. Depending on the closing velocities and the amount of thruster propellant available, an ASAT can even get multiple attempts to hit a target. An ASAT requires much more delta-v to reach the target, but the job of actually hitting the target is easier.

Because chemical rockets are only efficient up to around 4km/s, there's a strong incentive to design an ABM missile without orbital capability. This is good enough to take out ballistic missiles because you can always count on ballistic missiles coming back down.

Such a weapon could be used against satellites in low Earth orbit, but the enemy will quickly adapt by launching his spy satellites into a higher orbit instead. The marginal costs of sending a spy satellite into a higher orbit are rather modest, considering how much delta-v is required just to get it into orbit at all. The extra costs of the higher orbit and more powerful optics are mitigated by the fact that these higher orbiting satellites now have a better field of view.

With higher orbiting satellites, ASAT missiles would need to have orbit capability. An arms race between ASAT launchers and spy satellite launchers is thus a conceivable scenario for bulking up orbital launch capability.

That said, I would be against it. The main appeal of spy satellites is that they can overfly other countries during peacetime without starting a war. But during a shooting war, UAVs provide continuous coverage which isn't blocked by clouds, and they're a lot cheaper. If an enemy shoots down some spy satellites, I'd bet the response would be to replace them with spy UAVs rather than more spy satellites.

Of course, this is all assuming the use of missiles rather than lasers. At the rate solid state laser technology is progressing, I think it's entirely possible they will become the de facto dominant ASAT weapon within a few decades. Research into high energy lasers has traditionally been promoted as a missile/rocket defense, but the unspoken fact is that these lasers will be able to shoot down satellites, as satellites are easier targets than missiles/rockets. While nobody wants to talk about it, it wouldn't take much to point a high energy laser into the sky to damage a satellite.

In this case, many countries could end up possessing potent de facto ASAT capabilities without much--if any--orbital launch capability. This is a contest which military spy satellites probably just lose. At that point, cheap expendable UAVs become the only sustainable option. Spy satellites can still be useful for long term missions during peacetime, but they'll be the first things to go if a shooting war starts.

Isaac Kuo



Analogies can be drawn from history, though you have to be careful. Sometimes not all the constraints are the same. For instance, examining the Naval history from World War I to World War II and reasoning by analogy into interplanetary combat, one might come to the conclusion that space war will lead to the development of a one-man fighter. But there are different constraints that will probably prevent his.

Having said that, examining Naval history might be illuminating.

Before the 1860s, the Battleship was the queen of the ocean. It had titanic guns capable of blowing enemy ships out of the water, and armor thick enough to bounce off enemy shells. Granted it had all the speed and turning radius of a pregnant hippo, but that didn't matter.

Until some clown invented the Torpedo Boat. These little gnats could run rings around the battleships, were too agile to be targeted by the battleship's guns, and had torpedoes quite capable of sending the battleship to Davy Jone's Locker. Especially since the torpedo boats would attack in packs of twenty or more. The battleship was much too ponderous to avoid the swarm of torpedoes the pack would launch.

So the Destroyer was invented. This name was actually short for "Torpedo-boat Destroyer." This was a speedy, agile warship with quick guns designed to chew up torpedo boats. Of course this ability came at a price. The destroyer speed came at the cost of no armor, and the quick guns meant they are too light to damage anything heavier than a torpedo boat.

The upshot of this is that destroyers are pathetically vulnerable to enemy battleships.

So destroyers and battleships have to support each other. Destroyers protect their sister battleships from enemy torpedo boats, and battleships protect their sister destroyers from enemy battleships.

What happens if you design a warship that is equally balanced with regards to armor, guns, and speed? You get a Cruiser. Since cruisers are not specialized, they are viable enough to operate independently. They can be detached from a fleet as a task force of one for missions such as convoy raiding, deep scouting, and related missions. Generally a cruiser can outrun anything it cannot outfight. Heavy cruisers have large endurance for long distance scouting. Medium cruisers are often used as raiders, on convoys and other soft targets. Light cruisers generally operate with a fleet, scouting and repelling attack by enemy cruisers and destroyers.

And as an aside, it really annoys the Nifflheim out of me (and Jim Cambias agrees) when so many science fiction authors mistakenly use the term "Destroyer" for the largest class of warship. As you can see above, "Destroyers" are the weakest classes of warship, short of a torpedo boat. This mistake happens in the otherwise excellent TV show Babylon 5, the otherwise excellent novel MY ENEMY MY ALLY by Diane Duane, and the, er, ah, Star Wars movies. Mr. Cambias is of the opinion that this is due to the perception that the word "battleship" is old and corny and the term "destroyer" sounds really awesome.

Go to The Tough Guide to the Known Galaxy and read the entries "BATTLE CRUISER", "BATTLESHIP", "BATTLE STATION", "COMBAT SPACECRAFT", "DESTROYER", "FRIGATE" and "SPACE FIGHTERS".

For some of the difference between wet-navy ship classifications and possible space-navy classifications, refer to Mr. Ruppe's analysis above.

One of the problems with figuring out how ships are going to fight in space (assuming that we have ships in space, which isn't as likely as I wish; and, that we're still fighting when we get there, which is unfortunately more probable) is that there are a lot of maritime models to choose from.

It's also true that some of the maritime models came from very specialized sets of circumstances; and a few of them weren't particularly good ideas even in their own time.

And it's also true that some of the writers applying the models have a better grasp of the essentials than others.

From Space Dreadnoughts edited by David Drake

Trade Offs

There is a trade off between armor, guns, and speed. Each comes at the expense of another. One method of displaying this is by a triangular graph. It has three scales for three variables. At any point on the graph, the percentages of each variable add up to 100%. There are areas of the graph. All points to the right of the red line have a higher percentage of weapons than of defenses. All points below the blue line have more weapons than propulsion. So the pointy bit that is below the blue line and to the right of the red line are all the points where the ship has more weapons than either propulsion or defenses.

This "weapon dominance" area is divided by the green line. All points above the green line have more propulsion than defense. So the blue area containing the letter "A" is the area where a ship will have more weapons than propulsion and more propulsion than defense. Indeed, in the adjacent purple area there are no defenses at all.

Additional Attributes

As an aside, Ken Burnside points out that there are actually five major dimensions of ship design: armor, guns, speed, endurance (how long between refueling and re provisioning), and command & control (how large the bridge crew is, which boils down to how many different tasks can be done simultaneously). He notes that if you just look at the first three variables, one would make the erroneous prediction that the battle of Jutland would have been an overwhelming advantage to the Germans task force. In reality, the British had the advantage because they built their ships with the endurance for long cruises and the Germans built their ships with an endurance of only two weeks.

The classic historical example is the Battle of Jutland.

The (British) Royal Navy designed their ships for long cruises and extended habitability. The result of this was that their ships could stay on station and cruise for longer periods of time. A lot of resources were put into habitability, and when you're out in the South Seas and something breaks, you pretty much have to fix it from parts at hand.

The (German) North Seas Fleet was built with more powerful boilers. It had more armor; its guns were built with a fairly precise targeting systems. It could remain at sea for 2-3 weeks, tops. If something broke, it was meant to go back to port to be fixed.

I've given the impression that a "balanced ship" should have 20% maneuver, 20% offense, 20% defence, 20% command & control and 20% strategic endurance, I apologize for the confusion.

I typically see it as 27.5% maneuver, 27.5% defense, 27.5% offense, 15% strategic endurance and 2.5% command & control.

Some ships will trade offense or defense for greater cruise endurance (RN model) and others will trade cruise endurance for maneuver and offense (High Seas Fleet).

It's harder to find examples of ships that trade C&C capabilities for any of the other four categories; you can find science fiction examples (and limited real world examples) of ships that trade off offensive capabilities for better C&C. (This is, as far as I know, one of the things that distinguishes a CG {guided missile cruiser} from a DDG {guided missile destroyer} - they're built on nearly identical hulls, but the CG has less firepower and more "keep track of what's going on" gear.)



This is the results of my playing around with allocating WWII ship classes on the grid. Be warned that the above classifications are totally my own invention, and are a gross simplification. Any actual Naval scholar will severely hurt themselves laughing upon viewing this. You are encouraged to make your own grid, incorporating the technological assumptions and limitations of your own SF universe.

Since making the above chart, it occurs to me that a ship ship with 50% weapons and 50% propulsion (currently marked as "missile") is a good description of an interceptor. "Long-range" interceptors are larger, have more endurance, but lower speed. "Short-range" interceptors have shorter range but a much quicker response time.

And the area marked "courier" can also be "fast scoutships", faster than the other scouts because they are totally unarmed.

In reality, when mapping existing wet-navy ships onto the graph, there will be some holes. There are certain classes of ship that are theoretically possible to build, but in reality would have no well-defined function.

For instance, I used the term "packet" to mean an armed transport (because that is how the term was used in the old Triplanetary board game). They are in the dark orange and neon green sections. In the modern wet navy, there ain't no such class of ship.

CDR Beausabre says the only use he can think of for such a ship in a science-fictional setting would be some kind of raiding ship, i.e., some sort of vessel designed for planetary raiding as an independent mission - strong enough to punch through planetary defenses, land and hold a perimeter to awhile, and then escape. Which sounds like the Nemesis from the H. Beam Piper classic SPACE VIKING.

Marko Karonen points out that packets did exist, but you have to go back to the Age of Sail to find them. They only had cargo space enough for VIPs and mail, which was of critical importance before the invention of telegraphs and wireless radio. This would make sense in a science fiction universe which lacked faster-than-light radio. Age-of-Sail packets had some weapons to defend themselves against small enemy cruisers, and to make them too costly targets for pirates.

Actually, that is the main reason to make a chart like this, to find the interesting holes.

When Dimitri Mendeleev invented the periodic table of the elements, there were interesting holes in it. Mendeleev made the bold statement that these holes represented elements that had not been discovered yet, and predicted their approximate properties by analogy with the surrounding elements. He was vindicated when a couple new elements were discovered, and matched the predictions. So when you make your own ship chart, you may find holes. Examining the type of ship that would fill the hole will have you think either: [a] "What a worthless class of ship." or [b] "Wait a minute! That sort of ship could be useful." And some of the worthless holes might spark an idea later, say a specialized ship for a specialized mission, like the Brittania from Doc Smith's GALACTIC PATROL.

Note that the graph only classifies the ships by their relative proportion of the three components. It cannot distinguish between a mini-pocket battleship with six units of weapons, three units of armor, and one unit of propulsion and a cyclopean blot-out-the-sun battleship worthy of Darth Vader with 60,000 units of weapons, 30,000 units of armor, and 10,000 units of propulsion. Both will appear on the same spot on the graph. The light blue "A" section is labeled "torpedo boat" but some types of destroyers will fit in the same section. The difference is in the mass of the two ship types, which the graph doesn't handle.

It is better than nothing, but use it at your own risk.


In the current "wet" Navy, a "Fleet" is more of an organizational fiction rather than an actual entity. A group of ships belong to a fleet. But what is generally encountered at sea is a "Task Force." A few ships from a fleet are "detached" to form a task force charged with performing a specific mission. When the mission is completed, the ships of the task force are dissolved back into the fleet.

There are two classes of ships in a fleet: Main Units and Auxiliary Units.

Main units include Dreadnoughts (which were never an official type of unit but is included here as a tribute to E.E. "Doc" Smith, who spelled it "Dreadnaught"), Battleships, Battlecruisers, Heavy Cruisers, Light Cruisers, Escort Cruisers, Anti-aircraft ships, Destroyer Leaders, Destroyers, Submarines, Submarine Minelayers, Minelayers, Aircraft Carriers, and Aircraft.

Auxiliary units include Destroyer Tenders, Sub Tenders, Mine Sweepers, Aircraft Tenders, Fuel Ships (Oilers and Tankers), Supply (Logistics) Ships, Transports, Repair Ships, Hospital Ships, Colliers (missile supply ships), and Ammo ships.

There are ships that generally operate on their own, apart from any fleet. These are called Independent Units. They include Cruisers, Submarines, Gunboats, Torpedo Boats, Minelayers, Sub Chasers, Yachts, Aircraft, and assorted auxiliaries.

Don't sneer at the auxiliary units. An army marches on its stomach, and a rocket ship jets with its propellant tank. The old bromide is that amateurs study military tactics but professionals study logistics.

Going off of a very rough historical comparison to WW1 and earlier naval organizations try:

Squadron = More than 3 ships of same type/class/mission.

Flotilla = more than 1 Squadron operating independently under one commander.

Division = same as a Flotilla except operating as part of a Fleet.

Fleet = Multiple Divisions.

The logistical support ships, cargo, colliers, oilers, etc. usually operated to support the battle Fleet (Flotilla etc) and could be called a Division, Squadron, or Fleet Train. Some support vessels were never organized into units at all.

The US Navy still uses Squadrons, but formed units are generally called Battle Groups or Task Forces when operating alone, though they are still part of the Fleet.


When translating wet navy concepts to deep space, "continents" or the "mainland" are Planets, "coastal" is Planetary Orbit, "islands" are Asteroids, and "the high seas" are Deep Space. Instead of a "coast guard" you would have an Orbit Guard. There was an old class of coastal defense ships called "Monitors", these would be Orbital Fortresses.

Of course ever since the writers of classic Star Trek took the movie The Enemy Below and re-wrote it into Balance of Terror, everybody knows that Submarines = Ships with a Cloaking Device. The advantage of submarines is that they are very good at hiding, and can attack while hid. In interplanetary terms, this would require a science fictional level of stealth, since by the laws of physics as currently understood interplanetary stealth is more or less impossible (see the entry "CLOAKING DEVICE" in The Tough Guide to the Known Galaxy). For a good treatment of this theme, read PASSAGE AT ARMS by Glen Cook. Early non-nuclear submarines needed sub tenders for logistical support. Nuclear submarines do not need them. Sub minelayers can lay mines without the large escorts that a surface minelayer requires.

Ken Burnside had this analysis:

There's a decent functional space to discuss here.

Most navies really have three sizes of ship.

  • Small ships
  • Medium sized ships
  • Capital ships

Most navies have two roles that ships are designed for:

  • Independent patrol
  • Main battle fleet

Independent patrol sacrifices firepower (and sometimes protection) for cruise endurance and multi-mission capabilities.

Main battle fleet requires ships to be 'honed to the bone' - anything that doesn't make the ship more capabile in a fight is usualy a luxury.

History hasn't been kind to independent patrol capital ships. They're generally too expensive for the benefit they give the navy (something that eats independent cruisers for lunch and can do commerce raiding. Jackie Fisher's Battlecruisers in WWI and the German pocket battleships are two examples.

So this leaves:

  • Frigate (Small ship, independent patrol)
  • Destroyer (Small ship, main battle line)
  • Cruiser (Medium ship, independent patrol)
  • Armored Cruiser (Medium ship, battle line)
  • Battlecruiser (Capital ship, independent patrol)
  • Battleship (Capital ship, battle-line)

Within each role, you have specific missions, and you'll have different sizes of ships within each niche, depending on what specific navies did with their doctrines.

The frigate is the smallest thing that can be armed with guns capable of doing shore bombardment.

The destroyer may have less armament than a frigate; it's job is to shoot down threats to the bigger ships in the battle fleet.

The cruiser is a frigate that's generally got more armament, more armor, and more survivability. It usually has greater endurance.

The armored cruiser trades endurance for enough armor to maybe survive a hit from a capital ship's gun without being mission killed, and usually has the same number of guns as the cruiser with heavier throw weights.

The capital ship has Massive Firepower and the armor to stand up to it. Endurance is usually traded off somewhere.

Independent patrolMain battle fleet
Small sized shipsFrigateDestroyer
Medium sized shipsCruiserArmored Cruiser
Capital shipsBattlecruiserBattleship
Ken Burnside

CDR Beausabre had this analysis. The topic is the design of the interstellar battlefleet for a fictitious race of aliens called the Loroi.

The design of the ships of fleet are driven by requirements. Requirements are driven by mission and threat. The Royal Navy of the 18th century had one mission (control of the seas through destruction of the enemy fleet and blockade of his ports) and one threat (the muzzle loading black powder cannon). And it remained pretty much that way until the late 19th century. Battleships were for the destruction of the enemy fleet, and frigates (or later "cruisers") were to be where ever power was needed that didn't rate a battleship. Then newer threats showed up (the torpedo boat, the mine, the submarine) and new ships (the torpedo boat destroyer, the minesweeper, and the subchaser) appeared to counter those threats.

As those threats grew more sophisticated, and others appeared (aircraft), the design of ships and ship types changed to match. Aircraft carriers, and ships designed to counter the air threat (the Atlanta and Iowa classes) appeared. Existing ships adapted to the new threats, as destroyers became the primary defense against subs, and anti-aircraft batteries sprouted on every ship. Specialized amphibious ships were developed as the mission of projecting power ashore through troops grew more important.

After the war, the carrier was supreme, not the least because its aircraft could deliver nuclear weapons. In the US Navy the offensive mission centered totally on the carrier, and the various escorts (cruisers and destroyers) became almost purely defensive, putting up a barrier of guns, then missiles against air, and then missile attack.

(In contrast, the Soviets worked almost solely on carrier killers - cheap platforms with powerful missile armament).

For the US Navy of the future, things are changing again - the next generation "destroyer" - the DDX - will be optimized for deep attack missions against shore targets (and some of the capability will be usable for sea control). The next 'cruiser' will be a dedicated air defense platform. And a new class of ship, the "sea fighter" will take the fight close inshore, to deny the enemy the ability operate in shallow waters.

To think about what the Loroi fleet should look like, we should ask:

  • What is the mission of the Loroi fleet, strategically?
  • How does the fleet carry that mission out?
  • What are the threats to the ships?
  • How can those threats be countered?
CDR Beausabre

From The Napoleons of Eridanus (Les Grognards d'Éridan) by Pierre Barbet (1970). Decadent pacifist aliens from Epsilon Eridani are invaded. Desperate for military know-how, they kidnap a group of Napoleonic veterans fleeing Moscow in the winter of 1812. The head veteran uses Napoleonic analogies to handle alien military units.

As you have probably noticed, the ships placed at our disposal correspond reasonably well to the means utilized in armies on Earth. Fast and lightly armed corvettes are our chasseurs, the frigates are our hussars (light cavalry) and dragoons (medium cavalry). The (heavy) cavalry (Cuirassier) is replaced by heavily armored vessels, not as fast as the others. As for the artillery, it has been replaced by missile launchers. Finally, the light vedettes with short-ranged laser-disintegrators can be compared to the infantry. It was by making use of such equivalences that I planned the battle of Usk.

From The Napoleons of Eridanus (Les Grognards d'Éridan) by Pierre Barbet (1970)

Sample Space Carrier

Dean Ing has some interesting speculations on space warships. From "Vehicles for Future Wars", collected in DESTINIES vol. 1, no. 4, Aug-Sept '79, edited by James Baen.

But what of vehicles intended to fight in space? As colonies and mining outposts spread throughout our solar system, there may be military value in capturing or destroying far-flung settlements -- which means there'll be military value in intercepting such missions. The popular notion of space war today seems to follow the Dykstra images of movies and TV, where great whopping trillion-ton battleships direct fleets of parasite fighters (ed. note: Battlestar Galactica and Star Wars). The mother ship with its own little fleet makes lots of sense, but in sheer mass the parasites may account for much of the system, and battle craft in space may have meter-thick carapaces to withstand laser fire and nuke near-misses.

Let's consider a battle craft of reasonable size and a human crew, intended to absorb laser and projectile weapons as well as some hard radiation. We'll give it reactor-powered rockets, fed with pellets of solid fuel which is exhausted as vapor.

To begin with, the best shape for the battle craft might be an elongated torus; a tall, stretched-out doughnut. In the long hole down the middle we install a crew of two -- if that many -- weapons, communication gear, life support equipment, and all the other stuff that's most vulnerable to enemy weapons. This central cavity is then domed over at both ends, with airlocks at one end and weapon pods at the other. The crew stays in the very center where protection is maximized. The fuel pellets, comprising most of the craft's mass, occupy the main cavity of the torus, surrounding the vulnerable crew like so many tons of gravel. Why solid pellets? Because they'd be easier than fluids to recover in space after battle damage to the fuel tanks. The rocket engines are gimbaled on short arms around the waist of the torus, where they can impart spin, forward, or angular momentum, or thrust reversal. The whole craft would look like a squat cylinder twenty meters long by fifteen wide, with circular indentations at each end where the inner cavity closures meat the torus curvatures.

The battle craft doesn't seem very large but it could easily gross over 5,000 tons, fully fueled. If combat accelerations are to reach 5 g's with full tanks, the engines must produce far more thrust than anything available today. Do we go ahead and design engines producing 25,000 tons of thrust, or do we accept far less acceleration in hopes the enemy can't do any better? Or do we redesign the cylindrical crew section so that it can eject itself from the fuel torus for combat maneuvers? This trick -- separating the crew and weapons pod as a fighting unit while the fuel supply loiters off at a distance -- greatly improves the battle craft's performance. But it also mans the crew pod must link up again very soon with the torus to replenish its on-board fuel supply. And if the enemy zaps the fuel torus hard enough while the crew is absent, it may be a long trajectory home in cryogenic sleep.

Presuming that a fleet of the toroidal battle craft sets out on an interplanetary mission, the fleet might start out as a group of parasite ships attached to a mother ship. It's anybody's guess how the mother ship will be laid out, so let's make a guess for the critics to lambaste.

Our mother ship would be a pair of fat discs, each duplicating the other's repair functions in case one is damaged. The discs would be separated by three compression girders and kept in tension by a long central cable. To get a mental picture of the layout, take two biscuits and run a yard long thread through the center of each. Then make three columns from soda straws, each a yard long, and poke the straw ends into the biscuits near their edges. Now the biscuits are facing each other, a yard apart, pulled toward each other by the central thread and held apart by the straw columns. If you think of the biscuits as being a hundred meters in diameter with rocket engines poking away from the ends, you have a rough idea of the mother ship.

Clearly, the mother ship is two modules, upwards of a mile apart but linked by structural tension and compression members. The small battle craft might be attached to the compression girders for their long ride to battle, but if the mother ship must maneuver, their masses might pose unacceptable loads on the girders. Better by far if the parasites nestle in between the girders to grapple onto the tension cable. In this way, a fleet could embark from planetary orbit as a single system, separating into sortie elements near the end of the trip.

Since the total mass of all the battle craft is about equal to that of the unencumbered mother ship, the big ship can maneuver itself much more easily when the kids get off mama's back. The tactical advantages are that the system is redundant with fuel and repair elements; a nuke strike in space might destroy one end of the system without affecting the rest; and all elements become more flexible in their operational modes just when they need to be. Even if mother ships someday become as massive as moons, my guess is that they'll be made up of redundant elements and separated by lots of open space. Any hopelessly damaged elements can be discarded, or maybe kept and munched up for fuel mass.

Dean Ing