|
Post by someusername6 on Sept 10, 2018 5:03:56 GMT
The problem with the chemguns in the anti-laser role in my experience to date, beside lower velocity, is that they tend not to have as much ammo and are more accurate than my railguns (useful for drilling through armor, less so for fire at a mass of laser mirrors from 800Km). My problem when trying to send nanodrones against a laser star tends to be that they get fried before the projectiles arrive -- velocity matters maybe more than the score metric on this thread. If the problem is just accuracy, try shooting those 152 mg flak needle projectiles from Apophys? They blow up and make a mess all over the place, not just on a single spot (though that might not be as useful at lower velocities).
|
|
|
Post by treptoplax on Sept 10, 2018 12:08:01 GMT
The problem with the chemguns in the anti-laser role in my experience to date, beside lower velocity, is that they tend not to have as much ammo and are more accurate than my railguns (useful for drilling through armor, less so for fire at a mass of laser mirrors from 800Km). My problem when trying to send nanodrones against a laser star tends to be that they get fried before the projectiles arrive -- velocity matters maybe more than the score metric on this thread. If the problem is just accuracy, try shooting those 152 mg flak needle projectiles from Apophys? They blow up and make a mess all over the place, not just on a single spot (though that might not be as useful at lower velocities). Nice idea if you can make it work, but I've never been able to tune flak timing/distance parameters properly, and I tend to find that flak in large quantities produces massive lag. YMMV. I did some experiments with a couple of the smaller proposed guns last night and they died right away. Then I added a centimeter of PE armor to the turret and they took no meaningful damage closing from 1000Km to 800Km, and even with small numbers of drones some made it to 600Km, which is plenty of time for the first rounds to hit. It may also help that my drone design has a thick "radiation shield" of PE behind the gun - unless a chemgun magazine cooks off and takes it out the drone becomes a missile once the gun is disabled. I've been using jtyotjotjipaefvj's lovely (and cheap/lightweight!) Deep Cooker design to test against. The railgun proposed by someusername does quite nicely against it once laser-armored, even with the power input tuned all the way down to 15 Kw (!) leading to fire only once per second. By the time I added a turret and armor etc. etc. it was still about 17cm diameter, but that's noticeably smaller than the 25 I started with. Obviously such a design is hyper-optimized to take out near-stationary laserstars, but it does it wonderfully - a dozen 2Kc drones is enough. Going to be busy next few days, but I intend to experiment a bit further and post whatever I come up with late this coming weekend. Gunbarrels themselves seem never to be affected by lasers; not sure if this is a bug. There was IIRC a bug ages ago where they would block fire but not be damaged, don't know if this was resolved by just making them transparent...
|
|
|
Post by jtyotjotjipaefvj on Sept 10, 2018 14:36:55 GMT
Disregarding the silly point system, here's my entry: Comes out as 346 260 points. And if you're more serious about deleting laserstars (or any ship really), this should be good:
|
|
|
Post by treptoplax on Sept 10, 2018 15:02:30 GMT
Sorry for it only gets up to 1440000... I missed typed one more 0... [snipped small 5.5Km/s coilgun] That looks pretty good, actually! It does point a out a kind of limit in this scheme, though; my 202Kw drone is 23cm diameter. Even if you can make a magical 100Km/s 1 cm-diameter railgun somehow, the power supply will be big enough that the drone will be big anyway, or alternately it will fire once a minute.
|
|
|
Post by treptoplax on Sept 10, 2018 15:23:17 GMT
Disregarding the silly point system, here's my entry: [railgun!] Comes out as 346 260 points. And if you're more serious about deleting laserstars (or any ship really), this should be good: [yet more railguns!] Oh, wow. That's interesting - a good chunk smaller than any high-speed railguns I'd run across. Would definitely be a step up in size and cost but worth a good look. That actually has enough velocity I'd be tempted to try a MPD engine on it (maybe resistojets for attitude?) - put that in a needle-shaped-drone with a 10Mw reactor and let it sit 1000Km out shelling the enemy...
|
|
|
Post by someusername6 on Sept 10, 2018 15:36:58 GMT
It does point a out a kind of limit in this scheme, though; my 202Kw drone is 23cm diameter. Even if you can make a magical 100Km/s 1 cm-diameter railgun somehow, the power supply will be big enough that the drone will be big anyway, or alternately it will fire once a minute. I assume you are using a fission reactor for power -- you *could* just make it extremely thin and tall, if you don't mind a drone that is several meters tall and has only a few centimeters of diameter. But launching it from a ship might be a pain. I also think that a higher velocity railgun (50+ km/s) will do the job much faster, even without abusing the 1000 km range limit of the laser -- but it will be really hard to beat the current cost of a dozen drones at 2 kc each.
|
|
|
Post by treptoplax on Sept 16, 2018 20:39:03 GMT
Did some experiments, interesting results. One, asking for a super-tiny, super-high-speed railgun isn't that useful. Power required goes up with the square of velocity, so the power supply gets to be a limiting factor at high velocities anyway. I built a drone with the jtyotjotjipaefvj railgun (with MPD drive and resistojets for attitude control). It was pretty neat, but a) the AI couldn't figure out how to use it properly at all, b) it was just big enough that high-powered lasers seemed to be somewhat effective, and c) it cost 100x as much as the tiny drones I started with. Could be really useful with the right battle doctrine, maybe? Two, I underestimated small railguns. Chemguns really only make sense if you want >1g projectiles (and I'm not sure you do?), or you want extreme rates of fire for short periods. Once you have more than a few hundred rounds of ammo the small railguns are probably more cost effective, unless you're extremely power-limited for some reason. Third, don't always go for maximum accuracy. Tuning it down a bit helped a lot for some of these. Finally, I'm not sure there's a viable stopping point between tiny swarm drones and drones with capitol-ship-weapons. The Fgdfgfthgr coilgun was good (that long capacitor is awkward), but not really enough of an advantage over the slower railguns to be worth the cost/bother. His chemgun is devastating if it can get in range (good for a different style of drone), but can't easily have enough ammo to work as a long-range weapon. To my surprise the someusername6 gun with a few tweaks seemed to work the best despite the moderate velocity - it made for a really cheap drone with slightly higher performance and a far larger magazine then the chemguns. Rate of fire is lower (90 RPM), but that wasn't a huge deal. This drone is hilariously effective against MPD laserstars, and works better than you would think against anything else, given the cost. Even against opponents who can dodge, just spamming small waves of these to force them to expend deltaV might be a viable tactic.
|
|
|
Post by jtyotjotjipaefvj on Sept 17, 2018 9:49:43 GMT
Here's a thinner gun and a thinner drone, with higher muzzle velocity too. 10 of them can close to around 120 km of a deep cooker before the last guns fall off. You can probably beat any stationary long-range design with a few of these. They're a bit more expensive but who cares about ten kc here or there. Drone export: Needle.txt (5.1 KB) Screenies: Gun: It also gets around 1.1 million points. Drone:
|
|
|
Post by treptoplax on Sept 17, 2018 15:39:13 GMT
Here's a thinner gun and a thinner drone, with higher muzzle velocity too. 10 of them can close to around 120 km of a deep cooker before the last guns fall off. You can probably beat any stationary long-range design with a few of these. They're a bit more expensive but who cares about ten kc here or there. Drone export: [snipped pics] It also gets around 1.1 million points. Drone: [snipped pic] Probably not going to get a chance to play with that for a few days, looking forward to it. Aspect ratio seems a bit extreme but that's pretty much what I was asking for, I guess...? I'm actually surprised a Deep Cooker can survive long enough for those to close within 120km (unless you specifically ordered them not to fire to see how close they could get). I would suggest thinning out the barrel armor to reduce accuracy - with a bit of shotgun effect those should easily be able to disable stationary targets (any exposed lasers or radiators, anyway) nearly before they're in effective laser or railgun range at all. Obviously this will reduce ability to cut through armor, but if you can strip the lasers from long range it should easily be able to coast into medium range to finish with more concentrated fire. The 120km number is especially interesting. (Haven't properly tested or really worked the numbers carefully on this, but speculating). If it can close to 120Km on a Deep Cooker it can close to about that distance on pretty much anything. At 120Km time-to-target for that gun is 17 seconds(ish). To change projected position 50m in that time requires acceleration of about 300mG, consuming maybe 6m/s of dV. Hardly anything can sustain that for more than 10-15 minutes?? I'd really like a better understanding of long-range evasion in general. Naively it seems like dodge distance ought to be 1/2at^2, so proportional to acceleration and square of time-to-target, but if you're avoiding a continuous stream of fire and have to vary randomly I wonder if a random-walk square-root factor comes in and it ends up back proportional to time...
|
|
|
Post by jtyotjotjipaefvj on Sept 17, 2018 16:54:49 GMT
]Probably not going to get a chance to play with that for a few days, looking forward to it. Aspect ratio seems a bit extreme but that's pretty much what I was asking for, I guess...? Against lasers, aspect ratio doesn't make any difference. Diameter and armor thickness are the only two metrics relevant to survival, if you discount maneuvering. The first shots did land at around 450 km, but for some reason the cooker decided it doesn't care about getting shot. I've seen it do this before as well: The last two lasers ended up killing all the drones once they got closer. I don't really see the point of that. If you want to shotgun lasers, you can just target all the lasers. Or leave the guns to fire untargeted, which means they'll cover the entire surface of the ship, but will never waste ammo to missing, which would be an issue with a fixed spread angle. Plus the added barrel thickness may be relevant to keeping the gun cool.
|
|
|
Post by treptoplax on Sept 17, 2018 17:07:17 GMT
]Probably not going to get a chance to play with that for a few days, looking forward to it. Aspect ratio seems a bit extreme but that's pretty much what I was asking for, I guess...? Against lasers, aspect ratio doesn't make any difference. Diameter and armor thickness are the only two metrics relevant to survival, if you discount maneuvering. The first shots did land at around 450 km, but for some reason the cooker decided it doesn't care about getting shot. I've seen it do this before as well: [snipped pic] The last two lasers ended up killing all the drones once they got closer. I don't really see the point of that. If you want to shotgun lasers, you can just target all the lasers. Or leave the guns to fire untargeted, which means they'll cover the entire surface of the ship, but will never waste ammo to missing, which would be an issue with a fixed spread angle. Plus the added barrel thickness may be relevant to keeping the gun cool. Sorry, by aspect ratio I just meant the diameter/length, was unclear there. As to spread: yes, but you'ld have to switch targets while the first shots were still in flight if there are more targets than guns (plus the AI will never do that right - I tend to do mostly AI-v-AI testing out of laziness). The gun I was using had a much larger spread; the first volley didn't drill through the ship (!) like that one, but it did take out most of the lasers and the others and railguns didn't last long, which is all it needed to do - any shots that go through the ship are wasted! I'd bet with a spread of maybe .02 (or a bit less, I think? don't have design in front of me now) you'll find that as few as four or five drones is plenty to disable the target before it finishes them off.
|
|
|
Post by treptoplax on Sept 17, 2018 18:06:54 GMT
And to add briefly (rather than keep editing my last comment) - to a large extent having a larger spread is a workaround for limited capabilities of the targeting algorithm. In many cases it's bad enough that "'#$%# it, let's just fill their flightpath with a diffuse cloud of osmium dust" is a reasonably viable way to strip all the long-range lasers that are the only threat to drones so narrow anyway.
|
|
|
Post by thorneel on Sept 17, 2018 19:53:22 GMT
Here's a thinner gun and a thinner drone, with higher muzzle velocity too. 10 of them can close to around 120 km of a deep cooker before the last guns fall off. You can probably beat any stationary long-range design with a few of these. They're a bit more expensive but who cares about ten kc here or there. If you end up against projectiles, would it help to make those radiators wider and shorter (even if it doesn't look as good) so they make a smaller front target?
|
|
|
Post by treptoplax on Oct 1, 2018 3:45:40 GMT
I tried making jtyotjotjipaefvj 's coilgun less accurate, but as he predicted they just overheated immediately with less barrel armor. (We really need active cooling for barrels). I still think a little less accuracy is ideal in this role, if only due to limitation of the targeting. I tweaked my drones again a bit to up the rate-of-fire a bit - cost is up to just under 2Kc now. For what it's worth, lasers are mostly ineffective against them past 800Km or so and only kill them fast inside about 250Km. Five or ten of them will maul standard MPD laserstars: here's what ten of those attacking a Deep Cooker looks like: First shots: A moment later shots are denser; late shots have caught up because the drones were moving at a good fraction of muzzle velocity. One of about four pieces.
|
|
|
Post by airc777 on Oct 13, 2018 18:20:35 GMT
Well, I lose. But here are some high cyclic rate guns.
|
|