|
Post by blothorn on Oct 5, 2016 6:59:26 GMT
Hm. They seem to have accuracy problems--pitting 20 Hailstorms against four of your destroyers (since they are about a quarter the price/mass of my carrier) almost all missed. A total of one drone died (either to a missile or laser) before they all ran out of DV--by that point the drones had crippled one destroyer and moderately damaged two others (and the carrier had 180 Hailstorms left to send). Replacing the flak warhead with a 1.91 kt nuke was much more effective--they still struggled to get outright kills (thanks to its low-profile radiators, the Hailstorm is remarkably nuke-resistant), but quickly immobilized the drones. It increased the price by 21 credits, but look at that acceleration and dv: (You could probably get back the the old cost at lower mass/length by dropping some propellant.) (You could actually fit a 10kt warhead and still reduce mass relative to the frag version, but my cheapest 10kt warhead is 226c.) My only drone to have good success against the nuclear version was the Meteorite, which could open fire at 62.5km and thus got a good bit of damage in before the missiles arrived. This could have been remedied by launching the missiles on the strategic map and intercepting the drones away from the ships.
|
|
|
Post by blothorn on Oct 5, 2016 5:51:31 GMT
I launched 120 drones into a battle once and vowed never to do it again . I tend to use groups of 5-40 drones at a time to avoid both framerate and friendly-fire issues. It is also generally a good idea to avoid excessive clumps of drones tactically; if you have too many drones in the air at once it becomes cost-effective for your opponent to try to dodge them, and large groups are very vulnerable to nuclear missiles. Edit: speaking of miniature missiles, I keep meaning to see if it is useful as point defense.
|
|
|
Post by blothorn on Oct 5, 2016 3:26:12 GMT
My investigations into capital ship armor suggest that for the weight of the armor necessary to survive a broad variety of weapons, you can carry enough ordnance that you never need to see it. The carrier itself is lightly armored and armed--just a laser to give some breathing room if engaged by drones or missiles (although if that happens, something has gone terribly wrong). Pre-deploying a squadron of drones for added point-defense is highly recommended. The complement is 320 103kt missiles, 200 Hailstorm drones, and 80 Meteorite and Solar Flare drones. The Hailstorm is just a step up from a missile: very light and cheap. The kinetic cannon is devastating but short-ranged; it can make quick work of ships that have lost their lasers, but is unsuited to head-on assaults against balanced ships. The Meteorite and Solar Flare drones are more technical. The Solar Flare's laser makes it an excellent anti-drone option, while in a pinch it can turret-snipe against capital ships (although its effective range when doing so is short enough that going for the kill with Hailstorms is likely a better option). The Meteorite features a long-range (>100km) railgun; while its 1g projectiles stand little chance against a properly-armored hull, it can hope for lucky hits on turrets and radiators from very long ranges. This carrier relies on progressively stripping defenses: a missile can strip Whipple shielding, making the hull vulnerable to the Meteorite's railgun. Or the Meteorite can knock off lasers or radiators, giving the Hailstorm a chance to get close. Against enemy drones and missiles, the shear volume available to the carrier (and the high DV of the missiles and Hailstorm drones) allow it to trade 1:1 against most enemies. Note: Apologies for attaching the full UserDesigns; I am not going to attempt to isolate the transitive closure of all of this. Attachments:UserDesigns.txt (187.79 KB)
|
|
|
Post by blothorn on Oct 5, 2016 1:16:49 GMT
Note that many capships have <250mg of acceleration; the change in distance travelled over 10s due to acceleration is 490m/acceleration. So the gunship (190 meters, 170mg acceleration) can change its position by less than half its length (83 meters) by a full burn, and will expend 32.5 tons of propellant (0.5% if its total) to do so. (Granted, those numbers look a bit more favorable if its tanks are near empty, which they are likely to be in combat.) Turning may help a bit, but given slow turn rates and the considerable swing of the ship's stern (particularly on long ships with an armored citadel at the bow), not all that much.
To me, it looks like a fairly good tradeoff to burn ammo to force your opponent to maneuver.
|
|
|
Post by blothorn on Oct 4, 2016 16:53:06 GMT
Another: add a digit of precision to the laser cavity length (I think). It only has one, so the jump from 1eN to 2eN is very abrupt.
|
|
|
Post by blothorn on Oct 4, 2016 16:48:53 GMT
I believe that the developer has stated that railguns/coilguns are not using capacitative smoothing, and that power is supposed to be the peak power. So varying with rate of fire is not broken, but efficiency is off by a few more orders of magnitude than the trivial calculation suggests.
|
|
|
Post by blothorn on Oct 4, 2016 8:57:57 GMT
Note that there are a bunch of conditions that need to be tested fairly independently. 1g hypervelocity slugs are an extreme challenge for belt armor, but are fairly effectively broken up by Whipple shields. Expect your primary concerns for the main armor vs. them to be handling plasma heating and avoiding spalling vs high but distributed loads. Medium-mass, medium-speed projectiles are likely to come through the Whipple shield more intact, and so test armor performance against penetrators, while still being hypervelocity against many armors. Low-velocity projectiles are likely to actually ignore the Whipple shield and thus have the best concentration of force against the main armor, but will usually hit below the speed of sound in the armor (meaning that a very different set of characteristics are in play against them.
I am surprised I have not seen much mention of UHMWPE; it has the best tensile strength/mass ratio in the game, with decent shear modulus, relatively low Young's modulus, and excellent speed of sound. The relatively low shear modulus probably means that there are better options for resisting hypervelocity projectiles, and the low melting point means you want to keep it out of the way of plasma. However, its relative flexibility and high tensile strength:weight ratio should make it very resistant to spalling itself while being the best armor in the game against low-velocity impacts.
Applying a combination of some informal testing and my knowledge of the real-world physics: * Be careful with heterogenous layers in a single armor block: a stiff layer over a flexible layer will shatter before the flexible layer contributes much strength, and a flexible layer over a stiff layer will transmit more force than you expect to a backing. I suspect the first is why basalt over UHMWPE does not work as well as I hoped, and 50/50 combinations (hard outer surface over thick spall liner) did very poorly. On the other hand, against hypervelocity projectiles contribution from backing armor should be minimal anyway. * I have replicated the concerns about low-heat-capacity armor struggling with plasma heating. * I had noticeably better results from aluminum zinc magnesium than the comparable mass of aramid fiber (which is a relief, given their relative costs). * Maraging steel performed terribly relative to the more exotic materials (thickness adjusted for weight). On the other hand, it is likely to suffer less extensive cracking/spalling when close to its limit. * For its weight, 5cm of silica aerogel over an immense layer of UHMWPE took an amazing amount of abuse. The UHMWPE also cost 10x more than everything else on the ship.
I would encourage testing against the Gunship--the 286mm coilgun (10kg@5.14m/s) provides a very different sort of challenge that is not exercised by the railguns I am hearing about. (In particular, I doubt that 2mm of Whipple shield is optimized against something that heavy and that slow.
|
|
|
Post by blothorn on Oct 4, 2016 4:59:01 GMT
Hm. I have always been put off by the weight of armor, but I think I have also been building my ships too compact---larger ships make it less likely that stray shots will find radiators/modules.
But the impact on size is truly staggering; my 8km/s DV carrier went from a cross-section of 1830 to 5060, and had to increase total mass by 40% to achieve the same DV with its original armor (which was only 5mm aramid, 5cm amorphous carbon, 1cm UHMWPE to begin with). Granted, it was not AoN armored (it hopes to never see combat), but that is not much higher than what I am seeing people put on the non-citadel portions of "AoN" ships.
Silly question---what is the cost of hydrogen deuteride? It seems to have potential for comparable exhaust velocities while being slightly cheaper and 40% more compact, which should have a big payoff in tank/armor mass. Converting that carrier from hydrogen to hydrogen deuteride (stock engines), I reduced cross-section and cost by 40% and mass by 25%. Am I totally missing something?
|
|
|
Post by blothorn on Oct 4, 2016 4:25:34 GMT
I think assuming away orbital mechanics in battle is a *much* smaller problem than allowing drones to "ambush" enemies at 20km. The biggest lesson I have learned thus far is that real combat would take place at much higher ranges than this game seems to anticipate. Both very close opening ranges and "let me weather this fire until I can get into range" require an enormous suspension of disbelief--I expect that IRL, people would be opening with anything that can be fired sustainably (lasers and light-ammo railguns) even past 250km.
I think I would suggest: * Allow overriding the default range curve (particularly useful for missile "guns", or nuke/flak guns that do not require direct hits--my nuke coilguns maul drones at 50km, and have a suggested range of about 2km). * Make the AI use "ignore range" when tactically appropriate (not under ammo pressure, not taking power from weapons more effective at that range). * Never start battle under 50km (or even 100km; it is a compromise between realism and playability, not wanting to drag players through excessive waits when effective ranges legitimately are short). Right now, battles between missiles and ships without lasers start so close the ships do not even have time to find a preferred orientation. * Make the AI decide more intelligently whether to increase closing speed--right now it seems to always burn to close range when outside, which means that it burns too much fuel and winds up passing too fast to effectively engage when battles start at very long range.
|
|
|
Post by blothorn on Oct 4, 2016 2:29:08 GMT
Yes---it is essentially impossible to get missiles to follow a burn-coast-burn profile unless you are moving them as a group. Another case where pause/automation would be extremely useful, and the excuse "It just simulates the difficulty of keeping up with a battle in real-time" is ridiculous. There is no way the fleet admiral is pressing an "initiate terminal phase" button for every ship launched...
And even stronger proof that the fuses are broken: they do not detonate nuclear missiles before impact (which destroys the warhead "safely"). This is not often a factor because the guidance is so bad that they rarely actually hit, but using nuclear missiles against a station is surprisingly ineffective.
|
|
|
Post by blothorn on Oct 4, 2016 0:29:14 GMT
Note that in many circumstances, heavy Whipple shields actually make things *worse*, because more mass gets thrown at the belt armor. I suspect this was the problem with your 5cm aluminum/5cm RCC setup. The optimal Whipple shield is, I understand, proportional to projectile mass and inversely proportional to shell velocity (as higher velocity shells take less motivation to disintegrate). This also drives the usefulness of aerogel in a capship: the thin Whipple shields that are optimal against projectiles are susceptible to lasers, and a layer of aerogel can dramatically improve survivability while only minimally increasing shield mass.
I built a lot of short, tubby ships for better volumetric efficiency, but am coming over strongly to the side of long, skinny ships. Not only do thinner ships make a smaller target (even if total area is greater, area within the circle of fire of an enemy gun is smaller), but they are less susceptible to radiator loss (since radiators will be, on average, further from the aim point).
I also second observations that RCC is weak against optimized laser designs. I recently killed a laser frigate by cutting through its 5cm RCC armor to kill the reactors with a score of 2.5MW laser drones. If you want to leverage high conductivity and re-radiation, use diamond; if you want high heat capacity, use amorphous carbon; if you want low thermal conductivity, use aerogel (or basalt). That said, a smart opponent will be using lasers to snipe modules, not etch the Whipple shield.
Care to share UserDesigns so that we can test against this cruiser?
|
|
|
Post by blothorn on Oct 3, 2016 22:16:08 GMT
Yep. The AI needs to be more discriminating about its use of missiles vs. drones. (Proper course of action if someone intercepts nuclear missiles with a small number of drones should be to send a small number of missiles at a time, which will likely kill the drones while allowing the rest of the missiles to proceed to their targes.) And it should use "scatter" when drones are intercepted by missiles...
Missiles vs. drones gets even more fun with larger warheads; a single 10kt nuke detonated near the center of a stinger formation will usually kill them all.
|
|
|
Post by blothorn on Oct 2, 2016 21:21:28 GMT
Note that most stock missiles have very poor armor vs. lasers---missiles armored with silica aerogel seem near-immune to lasers unless you can get behind them and shoot the engine (and that with an optimized 100MW laser).
I have made some semi-effective kinetic CIWS--look for very high velocities (since missiles rarely mount Whipple shields or have any redundancy, the usual objections to hypervelocity projectiles are inapplicable) and fast turret traverses (since they will make most of their kills within 2-3 seconds within 10km, fast retargeting is key). The problem is that they do not scale well; all guns on the same ship target the same missile until it dies, which means that regardless of effectiveness each ship can only kill them at a rate of distance/muzzle velocity.
I suspect that railgun drones would be somewhat effective; each targets the closest missile, so they should divide their fire better.
|
|
|
Post by blothorn on Oct 2, 2016 20:59:07 GMT
Actually, that setup is completely broken. It is 8.5km/s right now, but 100% efficiency at 9MW would actually only be about 1km/s (7.71kg payload), even if I slowed the rate of fire down to 60RPM.
I really wish coilguns would get fixed; so few configurations conserve energy to within a few orders of magnitude that it is hard to make non-exploitative ones.
|
|
|
Post by blothorn on Oct 2, 2016 18:37:51 GMT
hahah , I'm actually also starting to get quite fond of diamond chambers ... other than that I'm playing with drones now... armed with micro nukes... one is firing a small 2.45kt missile and the other has a spinal conventional cannon firing the same warhead... its a bit of a FPS killer frankly when your drones spew 20 missiles each... also distracts the defensive lasers and the drones are free to pound with cannons... Yep. Something needs to be done about AI laser targeting priorities; payloads can be made a lot more laser resistant than drones, so it is almost always going to be better to kill of drones before trying to target their payloads. Right now just one of these can kill a handful of laser frigates (200 10kt nukes):
|
|