[Spaceships] Dealing with Cheap World-Killers?

[exalted]

Isn't it easy enough to simply say that the engine isn't 100% efficient like pretty much any other machine humanity have ever made. If it produce even a fraction of a percent in waste heat it gets more detectable and much easier to force to not go 100% all the time since it will simply melt the machine/space craft without rediculous radiators that would likely not be stealth treated.

[DataPacRat]

Originally Posted by malloyd

Whose light cones? Atoms that have been too close to the power plant?

I suppose creating energy means the local gravitational curvature goes up (there's more mass here than their was before) which you could probably handwave into some sort of consistent effect beyond slowing the expansion of the universe locally.

Welp, let me see...

Every particle is surrounded by a roughly spherical blob, a virtual line in space; anything beyond that line can never interact with that particle. Normally, that blob is around the distance of the Hubble horizon, a few dozen billion lightyears away. A more familiar such horizon is the event horizon of a black hole. Any time a particle accelerates in one direction, say to the left, then the horizon 'behind' it draws up more closely, so there's more space within the particle's horizon to the left than to the right. There are a couple of different ways to explain the next part - some involving the uncertainty principle, some involving virtual particles, etc - but with more open space to the left, there is more 'pressure' pushing the particle back to the right. Depending on which horizons are closest, that force is experienced as inertia or as gravity.

By using some clever trickery, machinery can be made that arranges for one of the very few particular sets of circumstances in which horizons can be consistently drawn closer on one side of the generator than on another, creating a long-term force towards that closer horizon; said force then being applied to any of several standard electricity-generating machines. Usually those circumstances involve extremely high accelerations of good numbers of extremely small particles; think more along the lines of 'strangely-shaped capacitors and vacuum tubes stuck inside MRI machines' than 'perpetual-motion bowling-balls and magnets'.

But that's mostly just background. (If any of the proposed tests for quantized inertia pan out, it may even be /true/ background, which is a bonus. :) But for plot purposes, it's mostly there to avoid the problem of relying on any "reverse the polarity of the neutron flow" sense-free technobabble.)


(Oh, one other possible solution to the world-killer problem has come to my mind: "those people who commit such acts rapidly find themselves removed from the genepool and noosphere", leading to a certain kind of evolutionary pressure. I might be able to handwave that once any initial batch of such people has been removed, the problem becomes the much smaller one of maintaining widespread sanity than converting an unknown number of unknown insane people. There are, of course, certain problems with this approach, but I'm offering it for any further inspiration it might, er, inspire.)

Originally Posted by exalted

Isn't it easy enough to simply say that the engine isn't 100% efficient like pretty much any other machine humanity have ever made. If it produce even a fraction of a percent in waste heat it gets more detectable and much easier to force to not go 100% all the time since it will simply melt the machine/space craft without rediculous radiators that would likely not be stealth treated.

Spaceships mentions that the 'Hot' reactionless drives give out such waste heat, while 'Standard' ones appear to be reasonably cool. (That said, I'm using the 'Exposed Radiators' design-switch as standard for this setting. Also the Slower Industrial Systems and Pyramid 34's armor-volume tweak, if they make any difference.)


(Edit: And I have some errands I can't put off longer; I'll prolly be back in a few hours.)

[Fred Brackin]

Originally Posted by DataPacRat

Let's see, SSp44, 12 (SM+12), -2 (Pyramid 34 armor), +5 (half-hour sweeps), -75 (7B miles), +32 (SM+31 array), +10 (in plain sight), +24 (silhouetted against deep space), -8 (stealth hull) = total -2 to detect from across the system. Even assuming a $900 quadrillion automated sensor array, it's not until the rock is 300 light-seconds away that it's autodetected, which doesn't give a lot of time to get any kind of interception into place.

Even with all the rear systems being Hot Singularity Drives, pushing at 6 gees, in 300 seconds, an interceptor could only move itself about 1600 miles, meaning they'd have to be peppered pretty thickly around any potential-target planet. And even then, a pile of rubble hitting a planet at .9c is almost as bad as a single solid rock doing the same.

How else might a Guard try to handle the situation?

If pretty much anyone can build a world-killer than pretty much anyone can build a defensive sphere of interceptors. Let's see how much damage you need to do to kill a world-killer.

So at SM+12 it has 300 dHP and 3300 would destroy it completely. It might still leave soem solid debris but I suspect that isn't a difficult amount of damage to reach.

Collision damage is 6Dx3 x the smaller unit's dHP x V where V is velocity in miles per second. 90% of lightspeed is about 168,000 miles per second. 63 x 20 x 168,000 is ......something more than 200 million pts of d-scale damage or perhaps more than 600x the amount of damage needed to take it to -10x dHP.

So even if this rapidly expanding ball of plasma does retain its' original amount of KE it's going to quickly be very widely dispersed. Chemical explosions produce particle velocities of thousands of meters per second. We should be looking at hundreds of thousands of meters per second at the least. Let's call it 500 kilometers per second. In just 10 seconds the plasma cloud will be larger than the Earth.

You would want the distance of interception to be as large as possible but you can use something smaller than the SM+4 or 5 that's the smallest Spaceships will let you build easily.

So you have a shell of interceptors around your target planet. Exactly how many and how far out is an engineering questyion but it realy ought to be doable.

[AlexanderHowl]

The problem is not necessarily taking out the 0.9c spacecraft. It is taking out the automated bombers that it launches at 10 AU from its target. At SM+12, the spaceship can carry 2400 SM+4 automated bombers, each which can carry 120 16cm bombs, meaning that they can deploy 288,000 16cm bombs. Each 16cm bomb masses 30 kg, meaning that they will have kinetic energy equal to 1 EJ each (265 megaton of TNT). Each automated bomber can also ram a target with a kinetic energy of 200 EJ (56 gigatons of TNT). The carrier never needs to approach within 1 AU to wreck the planet.

[Phoenix_Dragon]

You're setting up a scenario that pits extremely potent superscience power and propulsion, allowing for trivially easy relativistic attacks, against completely mundane sensor tech. Reliable defense is going to be extremely difficult because the scenario is so massively biased towards the offense. You're basically blindfolding a person, handing them a fist-sized rock, and telling them to block an incoming rifle bullet with it.

If you're going with GURPS Spaceship rules, the way you defend from it is... well, I guess you spend the thousands of quadrillions of dollars building a network of those far out enough from any inhabited planet to both detect and intercept a world-killer coming in through their sector of space (And the further out you put the stations, the smaller their covered sector of space!), and load it up with high-performance interceptors. Then you hope you get lucky. And while you're at it, you can hope your enemies don't spend a mere 0.001% of the money you spent on defense and send a hundred of these missiles per defensive station.

That, or you engage in some sort of extreme and absolute control over the technology, since every single person with a ship can turn it into a relativistic missile. Or require every single ship to adhere to approved flight plans, and launch kinetic-kill interceptors against any that deviate (With no FTL, I'm assuming these impactors came from somewhere in the system, which means they would have been easily observed leaving the system and trivially tracked from there on). Basically, you're not overcoming the huge advantage offense has, short of going full totalitarian and controlling everything. I wouldn't buy social engineering being the 100% success-rate that you'd need, unless you're literally turning people into mindless drones. If this tech disparity exists, I'd have a hard time buying that it hasn't been successfully used, and probably repeatedly.

You're literally handing every single ship pilot a WMD and trusting them not to use it. A fed-up pilot of some singularity-drive ship could turn their 100,000+ ton vehicle into a kilometer-per-second missile in just three minutes. That may not be a planet killer, but it's a city killer (A bit shy of Little Boy). Or more likely, an orbital killer, since the atmosphere would probably tear the ship apart before hitting the ground. And when you've got ships going around with infinite 0.5+ G acceleration, the difference between an optimal deceleration burn and a nuke-level impact is alarmingly small.

I'd add that part of the problem might be that the Spaceship detection rules seem rather pessimistic compared to reality (Possibly by several orders of magnitude). Unfortunately, even detecting it early may not help. Intercepting an object moving at 0.9c means you're trying to hit a target for which you only ever have outdated information; six seconds from passing, you know how it was maneuvering one minute ago. You get to guess. A second later, you've found out what it did over its last ten seconds. If you guessed wrong, it may already be outside your maneuver envelope. The only good news is that you might have multiple interceptors to make multiple guesses, and that massive relative speed means even one small interceptor should do the job. Just hope that world-killer isn't packing its own missiles, or you're probably losing just about every orbital structure in the system, too. And then the next absurdly cheap world-killer comes into the system unopposed.

[Fred Brackin]

Originally Posted by AlexanderHowl

The problem is not necessarily taking out the 0.9c spacecraft. It is taking out the automated bombers that it launches at 10 AU from its target. At SM+12, the spaceship can carry 2400 SM+4 automated bombers, each which can carry 120 16cm bombs, meaning that they can deploy 288,000 16cm bombs. .

In Spaceships cost scales with mass so for the price of that SM+12/100,000 ton world killer you could build 10,000 SM+4 automated interceptors. So the defense has a 4 to 1 advantage over the offense.

[DataPacRat]

Originally Posted by Fred Brackin

If pretty much anyone can build a world-killer than pretty much anyone can build a defensive sphere of interceptors. Let's see how much damage you need to do to kill a world-killer.

So at SM+12 it has 300 dHP and 3300 would destroy it completely. It might still leave soem solid debris but I suspect that isn't a difficult amount of damage to reach.

Collision damage is 6Dx3 x the smaller unit's dHP x V where V is velocity in miles per second. 90% of lightspeed is about 168,000 miles per second. 63 x 20 x 168,000 is ......something more than 200 million pts of d-scale damage or perhaps more than 600x the amount of damage needed to take it to -10x dHP.

So even if this rapidly expanding ball of plasma does retain its' original amount of KE it's going to quickly be very widely dispersed. Chemical explosions produce particle velocities of thousands of meters per second. We should be looking at hundreds of thousands of meters per second at the least. Let's call it 500 kilometers per second. In just 10 seconds the plasma cloud will be larger than the Earth.

You would want the distance of interception to be as large as possible but you can use something smaller than the SM+4 or 5 that's the smallest Spaceships will let you build easily.

So you have a shell of interceptors around your target planet. Exactly how many and how far out is an engineering questyion but it realy ought to be doable.

It occurs to me that the interceptors might be even cheaper than SM+4 craft. 16cm missiles cause 6dx4 * relative velocity (in mps) damage; if the missile is near stationary, and its target is at .9c, that works out to about 4dx1,000,000 damage. At under $7k each, it might actually be affordable to saturate near-planetary space with the things densely enough to turn the world-killer from an impactor to a rapidly-expanding cloud. (And if this particular scenario can be dealt with, then there's hope of dealing with related scenarios, such as the one below. :) )

Originally Posted by AlexanderHowl

The problem is not necessarily taking out the 0.9c spacecraft. It is taking out the automated bombers that it launches at 10 AU from its target. At SM+12, the spaceship can carry 2400 SM+4 automated bombers, each which can carry 120 16cm bombs, meaning that they can deploy 288,000 16cm bombs. Each 16cm bomb masses 30 kg, meaning that they will have kinetic energy equal to 1 EJ each (265 megaton of TNT). Each automated bomber can also ram a target with a kinetic energy of 200 EJ (56 gigatons of TNT). The carrier never needs to approach within 1 AU to wreck the planet.

Well, the first scenario I was pondering was the one requiring the least amount of investment by the attacker: "under a billion dollars" provides some high-useful mental shortcuts. (And after all, if an attacker /can/ flatten their enemy for that small an investment, why would they bother spending more?) It's also a near archetypical asymmetric warfare problem, as arranging for a defence appears to require significantly more investment than is required for the attack.

Originally Posted by Phoenix_Dragon

You're setting up a scenario that pits extremely potent superscience power and propulsion, allowing for trivially easy relativistic attacks, against completely mundane sensor tech.

Well, I'm open to any TL10 sensor tricks that come to mind. Eg, spreading a circle of sensor platforms along Jupiter's orbit to create some kind of multi-AU-baseline interferometer. Anyone have any GURPS stats for such a beastie, eg estimates of cost and/or Sensor rating?

Originally Posted by Phoenix_Dragon

Reliable defense is going to be extremely difficult because the scenario is so massively biased towards the offense. You're basically blindfolding a person, handing them a fist-sized rock, and telling them to block an incoming rifle bullet with it.

If you're going with GURPS Spaceship rules, the way you defend from it is... well, I guess you spend the thousands of quadrillions of dollars building a network of those far out enough from any inhabited planet to both detect and intercept a world-killer coming in through their sector of space (And the further out you put the stations, the smaller their covered sector of space!), and load it up with high-performance interceptors. Then you hope you get lucky. And while you're at it, you can hope your enemies don't spend a mere 0.001% of the money you spent on defense and send a hundred of these missiles per defensive station.

Well, yes, but let's focus on solving at least one form of attack first. :)

Originally Posted by Phoenix_Dragon

That, or you engage in some sort of extreme and absolute control over the technology, since every single person with a ship can turn it into a relativistic missile. Or require every single ship to adhere to approved flight plans, and launch kinetic-kill interceptors against any that deviate

I'd rather not have basic physics hidden from the population; I'm open to ridiculously-heavily-enforced traffic rules.

Originally Posted by Phoenix_Dragon

(With no FTL, I'm assuming these impactors came from somewhere in the system, which means they would have been easily observed leaving the system and trivially tracked from there on).

There's no FTL, but I want a good deal of STL interstellar travel.

Originally Posted by Phoenix_Dragon

Basically, you're not overcoming the huge advantage offense has, short of going full totalitarian and controlling everything. I wouldn't buy social engineering being the 100% success-rate that you'd need, unless you're literally turning people into mindless drones. If this tech disparity exists, I'd have a hard time buying that it hasn't been successfully used, and probably repeatedly.

I'm open to world-killers having occasionally been used in the setting's past, before a workable set of countermeasures was worked out.

Originally Posted by Phoenix_Dragon

You're literally handing every single ship pilot a WMD and trusting them not to use it.

Yep. Tricky problem, isn't it?

Originally Posted by Phoenix_Dragon

I'd add that part of the problem might be that the Spaceship detection rules seem rather pessimistic compared to reality (Possibly by several orders of magnitude).

Are there any fixes for that, whether official, Pyramid, or purely fan-built?


(As a related aside, according to Spaceships, TL10^ Singularity Drives have a minimum size of SM+12, which is an option I'm likely to keep in place. An SM+11 ship with a couple of oversized drive-units would thus likely be the smallest craft with a reactionless drive, unless there's some rule that lets doubly-oversized systems be spread across multiple ship-sections.)

[Fred Brackin]

Originally Posted by DataPacRat

Well, I'm open to any TL10 sensor tricks that come to mind.

The TL9 Very Large Gravscanner on p.63 of UT can be built without Superscience. It speaks of detecting million ton spacecraft rather than your mere 100,000 tonner but the .9c might make up for that. :)

Then there's using two of them for large scale gravitic interferometry or similar techniques.

[Phoenix_Dragon]

Originally Posted by DataPacRat

Well, I'm open to any TL10 sensor tricks that come to mind. Eg, spreading a circle of sensor platforms along Jupiter's orbit to create some kind of multi-AU-baseline interferometer. Anyone have any GURPS stats for such a beastie, eg estimates of cost and/or Sensor rating?

I don't think there is any "trick" that will make up for that severe a disparity. I have no idea how well such a huge interferometer would work. While I suspect it would have substantial complications just from light-speed delay, even that doesn't really matter. Detecting it is only part of the problem. Even if you do make such a monster, or give a substantial boost to sensors, you still have the difficulty of actually hitting the incoming missile(s), not to mention all of the potential WMDs flying around your system.

Well, yes, but let's focus on solving at least one form of attack first. :)

I don't think there is a solution, short of introducing more super-science tech (And we're talking pretty serious stuff, like FTL sensors or powerful beam weapons that are highly effective out to hundreds of light seconds). Especially since there's STL interstellar travel. Since offense is so much cheaper than defense, it's trivial to overwhelm any possible defense and obliterate everything within a system. Worse yet, you can have them come in from any direction, making it impossible to tell who launched the attack, or even framing someone else for it! You can't even use MAD to discourage attacks, unless you're willing to hit everyone around you if someone attacks you. Which is great if you're trying to setup a post-apocalyptic campaign in space, but not so much if you want any kind of large-scale society remaining to play in.

STL interstellar travel also kills the more totalitarian solutions, such as strictly controlling all in-system traffic and shooting anyone who deviates. All you need is someone outside your direct control deciding to buck the rules, and the WMDs are loose. Information and equipment control becomes impossible in the grand scheme, and you're entirely down to terminal detection and interception, which you're heavily disadvantaged at.

Yep. Tricky problem, isn't it?

It's looking like an essentially unsolvable one, given the proposed scenario. Certainly I can't think of any that seem believable, much less one that's easier to accept than "people just don't do that because it would be bad for the plot."

Are there any fixes for that, whether official, Pyramid, or purely fan-built?

Not that I'm aware of. The specific case I looked into (Infrared telescope detection of the Space Shuttle) had math that suggests real sensors outperformed GURPS Spaceship ones, in this specific case, by about 3-4 orders of magnitude. How this maps to other cases, I have no idea. But I'm pretty sure even a small TL10 sensor array would detect a operating external pulsed plasma engine, propelled by literal nuclear explosions, a lot further out than 1 AU.

[DataPacRat]

Originally Posted by Fred Brackin

The TL9 Very Large Gravscanner on p.63 of UT can be built without Superscience. It speaks of detecting million ton spacecraft rather than your mere 100,000 tonner but the .9c might make up for that. :)

Then there's using two of them for large scale gravitic interferometry or similar techniques.

Possibly oddly, the sensor rules for spaceships make no mention of the target's speed affecting detection range.


Still - a TL10 gravscanner has +12 to detection, costs $500k, and weighs a half-ton (as much as a single SM+4 ship's system). If we assume that it scales similarly to a Spaceships sensor array, a SM+16-system version would cost $500B, and have a Scan bonus of +24. Since it seems unlikely that non-superscience cloaking would affect such a scanner, I figure that the SM+16 gravscanner would autodetect a SM+12 ship at 700 light-seconds out... twice as far as the SM+31 sensor array, at a millionth the cost. That actually seems to come within reasonable distance for a planetary defence budget.

For a first approximation, let's say that said anti-world-killer defence budget works out to $1T of hardware (plus, longer-term, the maintenance and upkeep of said hardware), and that a first approximation is $500B on the scanner and another $500B on 75,000 16cm missiles. Is that enough to handle the reference problem, of the 100k-ton rock coming in from at least 0.75 ly out at 0.9c? If so, how much would a potential attacker have to spend to break through?

Originally Posted by Phoenix_Dragon

I don't think there is any "trick" that will make up for that severe a disparity. I have no idea how well such a huge interferometer would work. While I suspect it would have substantial complications just from light-speed delay, even that doesn't really matter. Detecting it is only part of the problem. Even if you do make such a monster, or give a substantial boost to sensors, you still have the difficulty of actually hitting the incoming missile(s), not to mention all of the potential WMDs flying around your system.

It looks like the gravscanner might be just such a boost to sensors. At least, it turns the problem from requiring a nigh-impossibly-expensive solution to a tolerably-expensive one, even if said solution still needs further work.

Originally Posted by Phoenix_Dragon

I don't think there is a solution, short of introducing more super-science tech (And we're talking pretty serious stuff, like FTL sensors or powerful beam weapons that are highly effective out to hundreds of light seconds). Especially since there's STL interstellar travel. Since offense is so much cheaper than defense, it's trivial to overwhelm any possible defense and obliterate everything within a system. Worse yet, you can have them come in from any direction, making it impossible to tell who launched the attack, or even framing someone else for it! You can't even use MAD to discourage attacks, unless you're willing to hit everyone around you if someone attacks you. Which is great if you're trying to setup a post-apocalyptic campaign in space, but not so much if you want any kind of large-scale society remaining to play in.

If it was an easy problem, I'd have solved it before starting this thread. :)

Originally Posted by Phoenix_Dragon

STL interstellar travel also kills the more totalitarian solutions, such as strictly controlling all in-system traffic and shooting anyone who deviates. All you need is someone outside your direct control deciding to buck the rules, and the WMDs are loose. Information and equipment control becomes impossible in the grand scheme, and you're entirely down to terminal detection and interception, which you're heavily disadvantaged at.

All fair points.

Originally Posted by Phoenix_Dragon

It's looking like an essentially unsolvable one, given the proposed scenario. Certainly I can't think of any that seem believable, much less one that's easier to accept than "people just don't do that because it would be bad for the plot."

I'm willing to head in weird directions here. "Launch nukes if they're in the wrong lane" traffic-cops have already been suggested. Maybe the only individuals trusted to handle the drives are copies of an uploaded-mind that has been tested to ridiculous extremes in virtual environments, or there are some nigh-superhumanly good patent lawyers involved, or somebody claims to be holding the universe hostage through an extension of the quantum-immortality thought experiment, or the piloting test is rigged to weed out anyone with a chance of being unstable, or all of the above are all going at once in a great big mess that's even more confusing than the current-day internet.

Does that level of oddity open up any possible solutions that you'd previously ruled out?

Originally Posted by Phoenix_Dragon

Not that I'm aware of. The specific case I looked into (Infrared telescope detection of the Space Shuttle) had math that suggests real sensors outperformed GURPS Spaceship ones, in this specific case, by about 3-4 orders of magnitude. How this maps to other cases, I have no idea. But I'm pretty sure even a small TL10 sensor array would detect a operating external pulsed plasma engine, propelled by literal nuclear explosions, a lot further out than 1 AU.

There's always the "There Ain't No Stealth In Space" page - IIRC, a 90's-era shuttle's maneuvering thrusters are detectable as far out as Pluto.

(The technobabble I'm using implies that the drive we're dealing with is about as thermally detectable as an MRI being used for a physics experiment - that is, not very.)