|
Post by jtyotjotjipaefvj on Aug 17, 2017 17:58:46 GMT
I realized that there's no real reason to keep my NTRs as the tiny pushers they were. So I made myself a proper engine, with 1 GN thrust and 1,500 TWR, costing 2.5 Mc. This allowed me to make a ship that can actually dodge, since 40 Gs of acceleration allow you to avoid railguns even at fairly close range. Apparently the crew can survive a 600g acceleration as well, which is what these engines do when the tanks run dry. They must be well trained to be able to do that! Ship design: linkEngine design: linkMandatory youtube video showing some dodging action:
|
|
|
Post by jtyotjotjipaefvj on Aug 16, 2017 21:23:07 GMT
I can confirm that this happens all the time with >10km/s velocities. It seems like the high acceleration causes numeric issues with the simulation. I've gotten ships up to 100 km/s by spending less than 10 km/s dv on a single burn.
|
|
|
Post by jtyotjotjipaefvj on Aug 16, 2017 21:14:03 GMT
And obviously maxing out enrichment to keep the core as small and light as possible.
For tiny NTRs you might have to decrease both to get the engine's heat output small enough to work with a smaller injector.
|
|
|
Post by jtyotjotjipaefvj on Aug 16, 2017 16:14:22 GMT
I'm more familiar with efficient NTR design but combustion rockets shouldn't be that different. Here's an example engine I've cooked up, pretty close to the 1MN range you want. Here are some pointers I use when making a new engine. -Throat radius should be very small. You might get a better TWR by increasing throat radius a bit and thinning the chamber walls, but the optimum is usually on the lower end. Since nozzle size is directly related to throat radius, you'll want to keep it on the low end. -Injector should be made out of Potassium since it's the lightest material, and have a fairly low RPM and a large size. If you want a slimmer engine, you'll have to make the injector heavier since a narrow injector won't be as efficient. -Adding a gimbal is a tradeoff between engine width and efficiency. Three ungimballed engines are more efficient and behave better for missile guidance, but take more room than a single gimballed engine, and may have worse turn characteristics. -You'll generally want exit velocity as high as possible. This means a stoichiometric ratio of 1 on bipropellant engines. As for the nozzle shape, you can usually find a good balance between diminishing returns on the exit velocity and nozzle weight. -Chamber wall should be made from diamond, since it's strong, fairly light and cheap, and can take high temperatures. -For NTRs, U233 oxide is the cheapest fuel, control rods are generally best made from titanium diorbide (I think). It's a good neutron stopper and can take high temperatures. Your core and control rod masses should be as low as possible to fit into a smaller space. It's also cheaper and lighter that way, so it's a triple win. I can't think of much else right now, maybe others will have something to add.
|
|
|
Post by jtyotjotjipaefvj on Aug 15, 2017 18:06:25 GMT
I was finally able to find a good balance of size versus thrust versus efficiency for drones, missiles and other smaller ships. Ethane is also a very cheap propellant, which helps keep the missiles affordable. Engine design: linkFlak missile made with said engine: linkThe missile is fairly slim and short, but has a decent range and acceleration. These factors make it a nice, spammable sidearm for medium-weight ships.
|
|
|
Post by jtyotjotjipaefvj on Aug 15, 2017 17:54:04 GMT
I'd say my space train is a pretty good damage sponge. It took me a good few minutes to kill an AI-controlled one with another one of the same design. Even though the heavy railgun can punch pretty much straight through the armor, anything targetable is placed on top of empty space, meaning it won't disable the ship. It only died once the untargeted fire was able to drill through to the crew modules, which means that pretty much all of the armor on the ship was shredded to bits. You could still disable crew modules with a lucky hit from a KKV type of weapon, but the crew is split into two places on different sides of the ship, meaning you'd still have to spend a ridiculous amount of missiles to reliably disable the ship. Here's a youtube video showing the battle and an updated design schematic. Design link
|
|
|
Post by jtyotjotjipaefvj on Aug 14, 2017 19:46:43 GMT
I noticed that if a ship gets cut in two and both sides would be a viable ship with guns, power and crew, only one side will be alive and the other will die. This is an issue since the design I posted can get its nose cut off and lose pretty much all weapons even though the tail portion should be able to function just fine. Therefore I revamped it to a layout with crew modules in the center section only, and weapons and larger generators distributed throughout the nose and tail parts. This way the ship can lose both the nose and tail and still keep going. I also modified the weapon loadout a little, giving the drones heavier cannons and mounting some light nuke missiles on the tail in place of thrusters. This makes the ship both more survivable and more deadly in close combat. Updated blueprint: And youtube video showing the ship beat 7 vanilla ships at once without dying, or even losing any sections:
|
|
|
Post by jtyotjotjipaefvj on Aug 13, 2017 22:27:48 GMT
The previous test gave me the idea of designing a ship that's very hard to kill, but rather due to extreme redundancy than heavy armor. As a result, I arrived at this broadside design: It has three separate sections, each of which has some fuel, at least four gimballed thrusters, a generator, a crew module, some weapons and enough radiators to run everything in that section. This means that even if the ship would get cut into bits, the separate pieces should be operational if they're not too damaged. Thanks to the broadside profile, just penetrating a few shots through the nose won't kill the ship either. At worst, you're going to lose one of three crew compartments. This means that the ship can usually return fire even if most of its armor has been shot off and the hull is riddled with holes. I've found that the design is very hard to kill completely, although the heavy guns do come off rather quickly. I did a couple of tests against mixed ships, and found that after some intelligent gun placement that guides enemy fire away from the crew modules, it will take pretty much forever to die to conventional guns. Heavy nukes or flak missiles would probably fare better, though the armor is at least somewhat resistant against those. In a test against two gunboats, one gunskiff and a few ships of my own design, it managed to kill all but one gunship and one ship that has a 25-cm VCS plate protecting it from all sides. My ship lived through the combat and was able to float away after leaving enemy weapons range after a close approach. Below is a youtube video showing the full combat if you wish to see it.
|
|
|
Post by jtyotjotjipaefvj on Aug 13, 2017 19:44:01 GMT
Doing some fine-tuning on my cluster frag warheads. I'm not sure if it's more efficient but at least it looks neat. This is made by 30 missiles traveling at 22 km/s, each deploying roughly 40 kg of diamond fragments, with a cost of roughly 5 kc per missile. The fragments penetrate all the way through my heavily armored target, and the missiles themselves create big holes on the backside of the ship as well. I also made some bits go missing from a gunship. I think the first 5 missiles did this. The rest just sailed through the gap, as seen in the picture.
|
|
|
Post by jtyotjotjipaefvj on Aug 13, 2017 18:43:54 GMT
I managed to improve the penetration quite a bit by changing to cluster frag dispensers instead of the conventional gun. Now instead of each drone shooting 600 15-gram rounds all over the target over a fairly long period of time, they deploy 600 51-gram fragments focused in a very narrow pattern surrounding the impact site of the missile itself, with the cluster warheads being deployed roughly 0.25 seconds before impact. This gives better reliability, as the drones won't need to spend a long time expending their ammo bins, and gives better penetration as well since there's more fragment mass as well as a tighter pattern of hits. It also lags the game far less. With the drones being deployed from a bus, they have practically a 100% hit rate, as they already start on a collision course with the target due to the bus using homing guidance as well. There's less damage to the full armor plate, but I did manage to cut the target ship in half due to the more focused penetration: You can see the vertical cut patterns surrounding the bigger holes made by the drones impacting the ship at around 22 km/s. The shrapnel cuts go through all layers of the armor as well, so they'd do a good deal of damage to internals if some managed to survive the kinetic impact of the drone itself. This was with just 30 drones as well, worth about 150 kc in total, making these drones a very cheap counter to heavy armor.
|
|
|
Post by jtyotjotjipaefvj on Aug 13, 2017 16:37:45 GMT
I'm curious about how these drones do against some of the armor schemes I've been researching. If willing and interested, change the armor on the target to 5cm spider silk followed by (1M graphgel and 1.0cm platinum Whipple sheild)x3. It did fare a little bit better, but it still died within the first few seconds of the first bullets landing. At least the back armor isn't totally see-through now, although it does have a good few holes in it as well. Target dead about 5 seconds in to the salvo: Front armor after the full salvo: Internal view, with shots coming in from the right hand side. Only parts of the radiation shields survived, everything else is gone. Rear armor. It looks a bit better than with the VCS plate.
|
|
|
Post by jtyotjotjipaefvj on Aug 13, 2017 12:14:08 GMT
I made an improved version of the drones I posted previously. They have 15.8 km/s dv, and they reach it in a bit over 8 seconds. Each drone is armed with three miniatyrized versions of the vanilla 60mm cannon, with 15-gram rounds and 30 rounds per second fire rate. The muzzle velocity is fairly low at ~1.1 km/s but that's not very relevant when the drones are doing 15 km/s. The drones also double up as kinetic penetrators, since they have a tendendy of colliding their target. They make a very cheap and efficient weapon that can quickly take down targets from around 200 km range. Design link: linkNothing can really survive a salvo of the drones, since they chew through even 20 cm of VCS plate in seconds. They're also pretty much impossible to avoid due to their ridiculously high acceleration. Maybe some equally fast point defense nuke missile would be a viable counter against them, but even that is questionable since the drones would have time to fire a good amount of rounds before the nukes would hit. Here's a video of them making short work of my VCS plated station: And here's another one just for fun, this time starting at 20km range. The drones look far more impressive in this one, although they aren't nearly as effective at penetrating the armor since they don't have time to accelerate before they've already passed the target.
|
|
|
Post by jtyotjotjipaefvj on Aug 13, 2017 10:08:26 GMT
It's even harder than you think — silica aerogel actually looks like weird clear blue-white ice. The game renders transparent materials as black. It looks like it should just be bluish glass with a bit of inscattering added to it. You could probably do it in blender, though I haven't got any idea how it works.
|
|
|
Post by jtyotjotjipaefvj on Aug 10, 2017 21:40:11 GMT
I had a go with a slim design as well. I made a 14cm wide drone that has a 14mm conventional gun on it. That way they should be useful against more than just laser stars. The drones have 6 km/s dv, and are carried by a drone bus with 14 km/s, for a total of 20. Both designs have a ridiculous amount of acceleration as well. The bus can go from 0 to 14 km/s in 7 seconds, and the drones take 20 seconds after that to reach their cruising speed. This means that against targets other than laser stars, you can do a low-velocity intercept that doesn't break the game and accelerate at the start of combat. This high relative velocity to the target means that the guns can shred through pretty much anything, not just unarmored laser stars. They have enough ammo for a 20-second burst, so no reasonable amount of armor should be able to survive against them. drone: linkdrone bus: linkStart of combat against a single deep fryer dl.dropbox.com/s/rwqg77196xus7bf/20170810235919_1.jpg?dl=0Opening fire at ~500 km. I fired just a few round per drone to save on lag, but they could go on for a good while. dl.dropbox.com/s/si2l5wkosnpaohj/20170810235949_1.jpg?dl=0At ~50 km, the deep fryer is dead. Still a couple of functioning guns and a lot more operational drones left. dl.dropbox.com/s/hwwjkkxbroz1ied/20170811000022_1.jpg?dl=0After beating the deep fryer with roughly matched credit cost, I tested the design against my standard hard target with a 20 cm VCS plate, a spall liner and a whipple shield. It lost power within the first second of shots hitting it, and by the end of the salvo there was very little armor left on the ship. First rounds hitting, and I think one drone hit it as well. dl.dropbox.com/s/wv7amh5gl8mhtpc/20170811001304_1.jpg?dl=0Lots of tracers. dl.dropbox.com/s/zrhrfgn2uljp99b/20170811001532_1.jpg?dl=0After the salvo, armor isn't looking too good. dl.dropbox.com/s/25es7b6wq8t5kcw/20170811001634_1.jpg?dl=0Even the back side is riddled with exit holes. 40cm of VCS gone at this point. dl.dropbox.com/s/nms0yhr4g4geigt/20170811001631_1.jpg?dl=0
|
|
|
Post by jtyotjotjipaefvj on Aug 10, 2017 17:45:10 GMT
Hahaha Yes. I managed to eliminate the 2.5Mc Laser Jellyfish with about 700kc of starkiller missiles. What kind of armor layout did you use? The drones I used were far faster than yours and they'd usually die before getting under 100km.
|
|