[Spaceships] How should a large station... duck?

[johndallman]

Originally Posted by DataPacRat

I'm hoping to find some numbers that add up to let the large cylinder be sufficiently unpredictable to dodge any good-sized near-c impactor before that impactor can change course.

This is simple, in principle. The key question is how much delta-V can you generate at right angles to the impactor's trajectory, after you detect him?

If this is definitely more than he can generate in that time for changing his course to try to hit, by enough to cover the size of your cylinder and the impactor, you can force him to miss.

If you can't generate even enough to cover the sizes of the cylinder and the impactor, he'll hit if he was already on course to do so when you detected him.

If it's in between those values, it's a matter of life-or-death tactics in a brief period of time.

You do have an advantage in that relativistic time dilation operates to his disadvantage, both on his reaction time and on his thrust. However, he's going to be ready to react, and the cylinder won't be unless it has an automated defence system against this kind of threat, which is on-line all the time.

[DataPacRat]

Originally Posted by johndallman

This is simple, in principle. The key question is how much delta-V can you generate at right angles to the impactor's trajectory, after you detect him?

If this is definitely more than he can generate in that time for changing his course to try to hit, by enough to cover the size of your cylinder and the impactor, you can force him to miss.

If you can't generate even enough to cover the sizes of the cylinder and the impactor, he'll hit if he was already on course to do so when you detected him.

If it's in between those values, it's a matter of life-or-death tactics in a brief period of time.

You do have an advantage in that relativistic time dilation operates to his disadvantage, both on his reaction time and on his thrust. However, he's going to be ready to react, and the cylinder won't be unless it has an automated defence system against this kind of threat, which is on-line all the time.

So far, for a one-shot emergency dodge system, the best option I've found so far is to use some of the smaller-than-SM+34 systems on a laser rocket. A SM+32 rocket, with total automation, costs $2.5 quintillion (not all that much on top of the $366 quintillion pricetag so far), requires an external source of a yottajoule of laserbeams (for which we can presume a small dyson swarm close to the star), and provides 0.2 gees of thrust (any more, and some of the interior will start falling upwards during maneuvers). A SM+33 fuel tank of ablative plastic costs $1 quintillion (plus $15 quadrillion for the fuel), and provides a delta-v of 0.15 miles/second (which, at 0.2 G, happens over about 2 minutes.

Ignoring orbital mechanics, with a diameter of 350 km, the station will diverge from its original course by its own width after about half an hour. I've already figured out that the cylinder's sensor array can auto-detect a SM+15 projectile 7 AU (or 58 light-minutes) away, and larger projectiles even further out... so we just might have a viable dodge system against simple near-c impactors.

Of course, more complicated attacks would require more complicated defenses, but I'm pleased that the numbers work out for at least the simplest case. :)

(Of course, after using this engine, the station would have to send out robo-miner craft to whatever other bodies in the solar system have the raw mass to refill the tanks, and the lasers might need to refill their capacitors, so it would be some time before a second course-change could be made, so a second impactor might be able to do what the first one didn't.)

Originally Posted by AlexanderHowl

A SM+34 object should really be more than a million square kilometers, as that suggests a mass of 10,000 tons per square meter column. Atmosphere masses 11 tons per square meter at sea level, meaning that the rest of the mass would need to be something else. A kilometer of water masses 1,100 tons per square meter and a kilometer of soil masses around 2,750 tons per square meter. You would either need an excessively thick ecosystem or a larger object to account for the mass.

I was mostly going by Spaceships' "Open Spaces" numbers; 10 SM+34 Open Spaces have that much surface area. ... I rather suspect that the SpaceShips system isn't quite designed for tin-can habs of this type, and so the scaling is probably off, but I don't have the math chops to go outside the books here.

If we assume high automation

The goal of this station is to keep a viable population of sapients living inside an ecosystem for arbitrarily long periods of time, even if those sapients happen to be stuck in the local version of the Stone Age for a few million years here and there. (They probably won't, with the occasional insertion of a vatfacced copy of one of the mind-uploaded builders every so often to check on things, but the goal is to be prepared even for that.) I'm generally assuming total automation is in place for every system.

Its SM+33 tertiary VRF laser weapons would have an output of 300 PJ and would deal 3d×5,000 (2) damage each hit, allowing it to turn most relativistic projectiles into plasma moving 0.9c in the opposite direction.

... A fairly important point that I'd forgotten. :)

Originally Posted by David Johnston2

I feel like pointing out something obvious. If the enemy can pick out the cylinder from that kind of range, then the people who built the cylinder can definitely see the kind of energy expenditure it takes to accelerate a missile to relativistic speeds; something which would be much more obvious.

I've heard a few mentions that GURPS' sensor rules are a bit wonky, so I'm not on sure footing saying what can and can't be detected at interstellar ranges; which is why I'm focusing more on the cylinder's nearest light-month or so radius.

Originally Posted by AlexanderHowl

Well, a far enough starting point (like a light-year) and a slow enough acceleration (no more than 0.01g) might go undetected with the proper design, but it would be unlikely to get above 0.05c in a realistic setting.

Beamriders. Gotta love 'em.

Anyway, destroying such a treasure would be beyond insanity, as it is a prize beyond compare. If it is a TL12 civilization, its productive capacity would be phenomenal (a SM+33 nanofactory could produce $3 quintillion worth of products per hour), and it could produce a SM+15 battleship every second with enough materials.

Even just a SM+29 TL10 Robofac (costing around 1% of the original station cost estimate), using the Slower Industrialization toggle, should be up to producing $1.5 quadrillion per day, which is nothing to sneeze at for anyone who can figure out how to gain control of the place. You'd only get one of those SM+15 battleships every 13h20m instead of per second, but really, how many of those do you need at a time? :)



Say, does the SpaceShips system include any mention of laser-boosted lightsails, outside of SS7's laser rockets that use ablative plastic?

[AlexanderHowl]

SM+15 battleships are only a couple trillion, so I think you are low balling the production figures. Since the production figures in Spaceships 6 assumes daily rates rather than hourly rates, $3 quadrillion a day (nanofactories multiply by 20) equals 1,500 SM+15 battleships per day. The Star Forge has nothing on the level of production.

One of the odd things about such a station is that it could easily be mistaken as a metallic asteroid rubble pile when it is inactive due to its low density (it is the mass of a large asteroid like Euphrosyne 31). Without any anomalous energy readings, it might be ignored for decades or even centuries, as rubble piles are annoying to deal with, and its true nature might not be visible until someone goes to mine it. Heck, something like that could be floating in the Kuiper Belt and we would not know it until it went active.

[Anthony]

The general case of dodging is that you have to move by more than the diameter of your object, in less time than the object has to correct.

This will usually not be possible for a competently designed impactor, as it will likely have as good course correction as the target, except in the extreme relativistic range, where light speed lag causes problems for both the defender and the attacker.

At 0.99c, in the reference frame of the target, it takes light 500s to cross 1 AU, and it takes the impactor 495s, so you have only 5s warning. This generally makes detection not very viable. However, the impactor has the same problem: you probably don't know the target's position with more accuracy than 'orbiting that planet', so if you've got 5 km/s delta-v available for course correction, you need to know the target's position something like 1,250s before impact, which means you need to be able to detect the target (and distinguish it from possible decoys) at 250AU.