|
Post by jtyotjotjipaefvj on Nov 4, 2018 22:59:58 GMT
Ran a quick test that seems to confirm my previous theory. 10 guns, 33 ms reload, 23 ms firing time; only one gun at a time fires due to power not being sufficient: 20 ms reload, 1.1 ms firing time, all 29 guns fire with each gun having 1 GW draw and only having 1 GW mounted on the ship: This is big news, since this means non-capacitor coilguns are not made useless by their extremely short firing times. You can get significantly lighter gun loadouts with coilguns at 20 km/s muzzle velocities than you could with railguns, assuming you want an absolutely massive rate of fire. Plus you'll be firing needles for free without losing efficiency to having a separate armature.
|
|
|
Post by jtyotjotjipaefvj on Nov 4, 2018 22:39:16 GMT
That does not work since the gun fires every frame, and hence draws 50 MW constantly. If I drop reload speed by a factor of 10, I can support 10 guns with one 50 MW reactor, which is interesting. I never thought the game would take that into account, since having a faster loader than firing time allows you to shoot at maximum firerate without increasing your power draw, effectively allowing efficiencies above 100%. Are they not all firing? How do you tell? Maybe it actually takes firing time into account for coil guns? Looks like about half of your guns are firing, which would correspond roughly with a 33 ms simulation tick divided by the ~1.5 ms firing time. If that's really how this works, non-capacitor coilguns are far better than I've thought previously.
|
|
|
Post by jtyotjotjipaefvj on Nov 4, 2018 21:30:15 GMT
One 50 MW reactor could drive ten or more of those 50 MW non capacitor coilguns. The loaders only draw 10 KW of power, the guns don't draw 50 MW at all times because they aren't filling capacitors. You should be able to half the size of the reactor. Electric actuators should greatly reduce the power draw of aiming the turrets so they can all turn and shoot at the same time. That does not work since the gun fires every frame, and hence draws 50 MW constantly. If I drop reload speed by a factor of 10, I can support 10 guns with one 50 MW reactor, which is interesting. I never thought the game would take that into account, since having a faster loader than firing time allows you to shoot at maximum firerate without increasing your power draw, effectively allowing efficiencies above 100%.
|
|
|
Post by jtyotjotjipaefvj on Nov 4, 2018 20:07:45 GMT
I've been tinkering with my 1kg Autocannon design and applying some of the lessons learned from my ultralight ~500g total mass drone gun design. - Replaced the original chromium vanadium steel barrel with one made of high-grade carbon fiber epoxy composite to save weight.
- Switched barrel armor from diamond to boron nitride for improved heat transfer
- Changed the propellant from nitrocelluose to octogen to increase weapon efficiency.
End result reduced a 13 ton gun down to 3.64 tons.
Your gun has a lot of excess mass in the loader and momentum wheels. Two of these will be significantly lighter than one of yours, while having identical rate of fire: Coilguns are also quite good at lower velocities, and they naturally fire needle rounds which are far better at penetrating armor: Two guns with a reactor and radiators have 6x the fire rate for roughly twice the mass:
|
|
|
Post by jtyotjotjipaefvj on Oct 31, 2018 23:37:08 GMT
Now that is a glass cannon! I wanted a bit more armor than that. Armor doesn't do much when you're fighting at 10 Mm. The MPD is enough armor to protect you from kinetics, and lasers should handle everything else. I'd need to build lighter reactors before I could build a craft that light, my current most watts per kilogram is also my smallest reactor, my 10gw reactors are about 1/3rd as efficient. Cool stuff though. Apophys has you covered: childrenofadeadearth.boards.net/thread/1624/ae-catalog-standard-modulesFor reference, here's the reactor my laserstar used: The reactors seem to be very close to the global optimum, so I just use all AE reactors these days. It's not worth the trouble to try to get past those in efficiency.
|
|
|
Post by jtyotjotjipaefvj on Oct 31, 2018 20:15:35 GMT
How? Heres a screen shot of two ships, one with three 1gw lasers with 75cm apertures, the other with thirty 100mw lasers with 50cm apertures. Both sets of turrets armored with 5cm of amorphous carbon. The 100mw lasers killed twenty turrets in the time it took the 1gw lasers to kill 16. I don't know why your results are different than mine. My 1 GW lasers destroyed 107 turrets in the time that the 100 MW lasers destroyed
22 turrets. After the ship with 100 MW lasers lost about 80 turrets it turned its nose away. Regardless, its the maneuver did nothing but worsen its already bad situation. So the outcome was not much different. It is worth noting that the ship equipped with the 100 MW lasers is about 12.5% heavier, 30 meters longer, and requires double the crew. Moreover, the ship with 1 GW lasers has nearly 3 times the turrets it needs whereas the 100 MW laser ship only has 16.7% more turrets than it needs. (I could not fit any more)
Your laser stars seem to be quite far from optimal. For reference, here's one of mine with a similar design: There's a good few of those in this thread: childrenofadeadearth.boards.net/thread/3376/challenge-750-ton-wonders
|
|
|
Post by jtyotjotjipaefvj on Oct 18, 2018 22:44:35 GMT
For those thinking about the RCS route: One of the obvious main advantage of separate RCS is, normally, that you can reorient without turning on your main engine and wasting a lot of fuel. Unfortunately, the AI doesn't know how to do this. For some reason, when the AI decides it needs to reorient and thrust, it turns on the main engine whether it is needed or not, someone negating this advantage. This still leaves precision gains to be made, however. You can prevent this by having an ungimballed main engine. If it still keeps burning during turns, your damping parameters might be too high. I haven't had these issues with any RCS design at least.
|
|
|
Post by jtyotjotjipaefvj on Oct 16, 2018 22:30:33 GMT
Did some testing of my design against the DCA gunship. It has a lot better armor and a highly advanced point defence technology where my framerate drops to 10 seconds per frame as soon as the first nuke hits and the dozens of drop tanks rupture, engaging maximum lag. As a result I needed to destroy the ship in one quick salvo of nukes, since waiting for more waves to arrive after that takes several minutes. I used this highly elegant solution of firing a wave of 140 Frag missiles followed by 80 Strikers. The Frag missiles blow a huge hole through the rear of the gunship and hopefully destroy something important in the process: A bit later the nukes appear. I'm treated to the game apparently freezing while playing the bling bling sound notification occasionally. Eventually the screen refreshes and I can see the end result of all my nukes having done their work: It's a little disappointing for sure, but once we wait for the flash to fade, we see what's left of the gunship. A significant portion of the rear seems to have flown off somewhere else: Here's some more screenies from tests that looked more impressive but ended in the game freezing, or me not firing enough missiles to kill the gunship in one pass. More impressive nuke hits: Not that effective return fire from the piddly stock guns: Close combat carrier operations: Another angle, same situation:
|
|
|
Post by jtyotjotjipaefvj on Oct 16, 2018 3:19:39 GMT
Since all the stock guns are hot garbage, I used a slightly different approach. I disregarded any guns completely and went full missilespam: close combat edition. It packs 360 Flak Missiles and 450 Striker Nukes, both of which are fast enough to get on an intercept course within ten seconds of launching. This means I can fire massive missile salvos while doing a decently fast flyby of the gunships in combat. 1 km/s intercept speed seems to be a good balance of number of salvos you have time to fire and the amount of time you need to wait for missiles to fly along. It has enough armor to shrug off stupid amounts of stock gun bullets, and can survive flying through a gunship formation while having just launched a hundred Strikers with only some minor radiator damage. It still mounts some direct fire armaments, so all requirements are fulfilled. It can probably disable an arbitrary amount of gunships on the first pass, although the armor starts to give in around 8 gunships in with a slower intercept. I could maybe push it up to 10 but I don't feel like doing any micromanaging right now. I also used only half the mass budget and less than half of the credit budget, but I couldn't think of anything worthwhile to put in there. The missile spam makes my usual hundreds of 60mm guns approach a bit unfeasible as well, so I just went with two placeholder lasers. You can inflict some damage with them by blowing up all flak launchers on the enemy gunships at least. I ended up lowering gunship count to 6 due to lag and missile guidance reliability issues. Design: Gun ship.txt (1.64 KB) Combat screenies: Start of combat, launching first wave. Consists of 30 flaks mainly used to absorb lasing damage, and 100 Strikers behind them. The Strikers' job is to cause enough damage that most gunships won't be able to fire or move. The second wave consists of around 60 Flaks and 60 Strikers. The flaks will be effective once the first wave hits, resulting in the gunships being mostly stationary, and disabling their lasers. Third wave to finish the job. 100 Strikers and 60 Flaks this time. The gunships are firing back but they don't do anything: First wave hits, doing a decent bit of damage. Only two Gunships are intact enough to fire significant amounts after this. Second wave hits, only three gunships are left alive, and they're all drifting. Third wave hits while the carrier flies by. Only one gunship is still technically alive. Coming in for a second pass: And job done. 6 gunships down with no significant damage. I didn't test it in Vesta Overkill, since the amount of lag would be just unbearable. But I'm pretty sure nuking absolutely everything should work just fine there.
|
|
|
Post by jtyotjotjipaefvj on Oct 15, 2018 17:26:07 GMT
I haven't tested 3 motor RCS that much, I've usually gone with 4 to save on propellant. I don't think there's too much difference between the two though. If engine diameter limits your 3-engine setup, you could always place all three separately and keep the same diameter as with the 2x2 setup. For radial engines you can at best get self-intersection errors this way. nah:
|
|
|
Post by jtyotjotjipaefvj on Oct 15, 2018 15:00:34 GMT
Separating rotation and acceleration is the main benefit. If your missile can't turn without accelerating sideways, small corrections needed by highly precise missiles are basically impossible. If you need a tiny adjustment burn to your trajectory, the rotation burn will impart more acceleration to the missile than the correction burn you want to do. Keeping your RCS symmetric around the CoM on your missile's long axis prevents this, as the opposing rotation thrusters will cancel each others' acceleration. I was thinking of 2 sets of 3 vernier motors on opposite ends of the missile, one at 0o, the other at 180o.
A pair of motors with 120o angle between them will produce as much thrust as a single motor (albeit at 50% efficiency) so this setup should remain balanced and 3x clusters should help keep the missile narrow.
Alternatively use 4 pairs of motors for minimum possible diameter.
I haven't tested 3 motor RCS that much, I've usually gone with 4 to save on propellant. I don't think there's too much difference between the two though. If engine diameter limits your 3-engine setup, you could always place all three separately and keep the same diameter as with the 2x2 setup.
|
|
|
Post by jtyotjotjipaefvj on Oct 15, 2018 10:14:52 GMT
What are the differences between making effective guidance for typical gimballed missile and an RCS based one? Apart from ubiquitous vernier overshoot issue (which alone is a big 'but') RCS could in principle make missiles easier to guide by separating rotation and acceleration. Separating rotation and acceleration is the main benefit. If your missile can't turn without accelerating sideways, small corrections needed by highly precise missiles are basically impossible. If you need a tiny adjustment burn to your trajectory, the rotation burn will impart more acceleration to the missile than the correction burn you want to do. Keeping your RCS symmetric around the CoM on your missile's long axis prevents this, as the opposing rotation thrusters will cancel each others' acceleration.
|
|
|
Post by jtyotjotjipaefvj on Oct 14, 2018 14:21:28 GMT
Ok, this one preformed far better in combat, and I've achieved 3.99% efficiency. I think part of it was it just did not occur to me to start with the arc lamp radius at 1 mm. Currently using 60 cm aperture radius so I can have a lot of redundant turrets without heavily increasing the crafts mass. It ate missiles pretty good, didn't do to bad against the Vesta fleet either.
Is it worth it to have two sets of lasers on the same ship, one with a small aperture set to only target shots and one with a much larger aperture set to target ships?
Looks like your range is set to 35 km? You can crank it up to 400 without any worries with that intensity, maybe even the full 1000 km. I've thought about having separate lasers for PD and anti-ship duty as well, but never got around to testing it properly. I'm fairly sure it would benefit you though, so go ahead and give it a try.
|
|
|
Post by jtyotjotjipaefvj on Oct 13, 2018 8:34:06 GMT
It's almost certainly due to the simulation tick rate. I've been playing a bit of From the Depths recently, you tend to get similar issues in that game too with high velocity projectiles, and both games are built (seemingly) in Unity. CDE is not built on unity, it's written in C++ and seems to use its custom engine.
|
|
|
Post by jtyotjotjipaefvj on Oct 7, 2018 22:40:55 GMT
Made some improvements to my design by reducing turret size which allows me to mount thicker armor on each turret due to lower surface area, and doing the same thing to the central PE plate protecting the ship's hull from lasing. Now I can beat 8 wheels simultaneously without losing a single turret. Updated design: jt Laser Spider.txt (8.44 KB) Loads of wheels: Also 5 deep fryers simultaneously, resulting in 13 lost turrets:
|
|