|
Post by blothorn on Oct 12, 2016 18:45:11 GMT
Also, my experience is that ultra-high-velocity light-projectile railguns are far less damaging than their kinetic energy suggests, even against targets without Whipple shields. I would suggest increasing projectile weight, even if it costs velocity.
The problem with relying on point-blank interception, though, is that flak is fairly effective within about 10km.
|
|
|
Lasers
Oct 12, 2016 18:41:57 GMT
Post by blothorn on Oct 12, 2016 18:41:57 GMT
How does one actually armor drones against laser? I tried to armor my drone in cms of silicon aerogel and boron but they melted away pretty fast. ...Or maybe it's because my drones are going up against 100 megawatt laser. Either way, still needs to improve them. Are they punching through the drone's armor or sniping the weapons (which for conventional guns is a near-guaranteed total kill)? If the latter: + Put the weapon on a turret. In my experience, no barrel material lasts more than a few seconds against a high-powered laser. + Use a good thermal armor on the turret. In my tests, it took a 500MW laser with a 3.7m aparture (2290MW/m^2 at 240km) 220 seconds to burn through a turret with 10cm silica aerogel. The downside of silica aerogel on turrets is that you cannot back it with a kinetic armor like you can for hull armor, but if the drones are close enough to take kinetic fire they should be well within their own effective range. (Aramid fiber also has excellent thermal performance, but to me the cost is prohibitive.) And yes, I think that drones vs. laser capships strongly favors the drones. It makes sense: when weapon/armor balance favors armor, the optimal ship is large, as larger ships have a better surface area:volume ratio and can afford heavier armor. When effective armor is impractical, balance tends to favor redundancy, and little is more redundant than a swarm of drones. My testing on armor suggests that this game is strongly in the latter category. That said, most of the weapon/armor imbalance comes from nukes, kinetic weapons, and turret sniping. As cuddlefish notes, nuke and kinetic physics are severely broken, and turret sniping is so effective mostly because of the lack of composite turret armor.
|
|
|
Post by blothorn on Oct 10, 2016 19:54:49 GMT
Below 3km/s I think Whipple shields are not much of a concern--on the scales at which Whipple shields are designed to work, this is rather low. On the other hand, you need a fair bit of mass to impart enough energy to bother armor at these low speeds; in my armor tests, a 1g@13.1km/s (86kJ) railgun penetrated armor better than a 5g@2.83km/s (8.4kJ) conventional gun, despite presumably suffering much greater fragmentation from the Whipple shield.
|
|
|
Post by blothorn on Oct 10, 2016 7:27:19 GMT
But note the lack of anything at the very small end--only 7 under 100kg, and none above 1kt/kg, while larger CoaDE nukes (such as the one above) can easily pass 20 kt/kg. (The lightest real 4Mt nukes are over 1000kg.) The mass/yield relationship is not linear--for a given mass, we are getting an order of magnitude better yield. Remember also (as noted elsewhere) that this compares boosted fission to the much more mass-efficient Teller-Ulam designs.
And the 20,000kg/m^3 fusion fuel densities seem absurd whatever the outcome.
|
|
|
Post by blothorn on Oct 10, 2016 4:41:16 GMT
The projectile was 3mm, I believe.
Whipple shields are weird--since they are dealing with intact projectiles at extreme velocities, they see the closest to Newtonian impacts; this means that areal density is about the only thing that matters. I experimented with a number of other Whipple shields but none decreased the thickness of main armor required (and I have suspicions, but no unambiguous evidence, that brittle materials such as amorphous carbon degrade faster. The RL papers I have found on the subject certainly support that idea). Alpha-2 titanium aluminide is the logical choice among ductile materials for its amazing specific heat (and fairly high melting point for a relatively light metal).
I would actually expect to see much more spalling from amorphous carbon; it is substantially less flexible than boron and has barely half the tensile strength. Note that the 13.1km/s projectile I was using is hypervelocity against both boron and amorphous carbon; it might be easier to distinguish between them with a projectile that is between their respective speeds of sound.
Note that for "flexible" you want low Young's modulus, not low shear modulus. Low shear modulus just means it is vulnerable to plugging (although more flexible materials are less vulnerable to plugging, so the low Young's modulus of spider silk compensates for its relatively low shear modulus).
|
|
|
Post by blothorn on Oct 9, 2016 21:30:28 GMT
Found a setup that will give a repeatable penetrates/does not penetrate verdict (without angling). Initial conclusions are that I have yet to discriminate between boron and amorphous carbon (which I find rather surprising, given their considerably different properties), and that versus 1g@13.1km/s projectiles, I have yet to find an advantage for Whipple shields heavier than 4mm alpha-2 titanium aluminide.
Lightest affordable armor thus far that stops that 1g@13.1km/s railgun is 4mm alpha-2 titanium aluminide, 1m gap, 9cm boron, 4cm spider silk. This also handles a conventional gun firing 5g osmium rounds.
Meanwhile, armor that does *not* stop a stock 286mm coilgun: 1cm alpha-2 titanium aluminide, 50cm gap, 5cm boron, 50cm gap, 30cm boron, 2cm osmium, 10cm spider silk. I am not seeing reason to doubt my earlier decision that capital ships are irremediably vulnerable, although a non-cheating version of that coilgun is getting rather large to fit on a drone.
|
|
|
Post by blothorn on Oct 9, 2016 19:18:48 GMT
I'm not sure how much 20 cm of aerogel can be considered feasible. It can be a fairly huge cost and volume investment, and might not last enough against a bigger laser setup That is actually the cheapest armor tested---aerogel is 5x the cost per unit mass of the boron, but this is 2/21 the mass of the other armors (aside from the ridiculously expensive aramid fiber). The volume is high, but still only adds 5% to the total radius of the turret. On this particular turret, it also adds 30% to the mass and doubles the cost. Otherwise--the laser I tested against is more powerful than almost all the lasers I have seen on the "laser" thread. The main point of these experiments was to test the hypothesis that it was optimal to leave lasers unarmored--one silica aerogel-armored laser lasting 40 seconds will easily survive long enough to destroy two lasers with negligible armor, while (in my case) costing the same, having 35% lower mass, and requiring half the power generation and radiators.
|
|
|
Post by blothorn on Oct 9, 2016 8:50:31 GMT
Nice! Interesting to see how large the difference vs. railguns of the 1cm vs. 3mm Whipple shield was; I would have expected the 3mm to suffice for such a light, high-velocity round. I would also experiment with some metal shields, as basalt is rather expensive for the purpose--I have had good results with 5mm aluminum zinc magnesium.
The one other thing I would keep in mind is cost; in my experience UHMWPE and the fibers make superb final layers (their relatively low Young's modulus allows them to absorb much more energy without fracturing than boron and the ceramics), but using nontrivial quantities is prohibitively expensive.
|
|
|
Post by blothorn on Oct 9, 2016 7:21:24 GMT
A few notes: + Gun-type plutonium weapons are impractical due to the slow assembly (although we seem to have the ability to isolate pure Pu-239, which might avoid the problem). Uranium is going to be much heavier. + The "more fissile material to reach a certain yield" is by a factor of about 10. + Gun-type weapons cannot scale up well, because each component cannot be critical on its own (and they cannot be extended in the direction of travel due to the slow assembly). + While several gun-type warheads were intended to detonate after impact, impact speeds were very low. Even a 3km/s impact is at about the speed of sound in uranium, which indicates substantial (and unpredictable) deformation which may interfere with assembly.
So yes, a gun-type weapon would give a (probably small) chance of post-impact detonation, but the small maximum yield and low efficiency suggest that a close proximity-fused implosion weapon would be much more effective.
|
|
|
Post by blothorn on Oct 8, 2016 22:21:04 GMT
Same setup, but now a somewhat larger turret--a 7.9m inner radius laser (the laser I was running these tests against, actually). Standard areal density is 210kg/m^2. Material | Time (1 turret) | Notes | 10cm amorphous carbon | 2s | I was expecting a bit more. | 1cm rhenium | 0s | Fairly reflective material with a much higher melting point than aluminum/silver. | 10cm boron | 0s | | 10cm aramid fiber | 65s | Only using 2/3 normal density because it is already ~10x the price of most of the others. | 20cm silica aerogel | 84s | | 5cm alpha-2 titanium aluminide | 3s | | 8.1cm silicon dioxide | 8s |
|
It certainly is feasible to give even huge targets protection from even very large lasers, and I expect that a laser ship with 10-20cm of silica aerogel would thrash a ship that did not, even accounting for the latter being able to afford more power or redundancy. However, there is a decision to be made: silica aerogel provides negligible kinetic protection, and aramid fiber is exorbitantly expensive in these quantities (the 10cm armor used here cost around 30 Mc on its own). I am increasingly of the opinion that not being able to put composite armor on turrets is substantially compromising the fidelity of the game balance--it forces a much harsher tradeoff between thermal and kinetic protection than I think justified.
|
|
|
Post by blothorn on Oct 8, 2016 8:30:44 GMT
AFAIK, in-game radiator effectiveness depends on surface area and thermal conductivity from the coolant pipes to the surface (plus some measure of coolant pipe size; I have found 2cm gives near-optimal results, although I am not certain whether it has an interaction with armor thickness). So dissipation/weight depends on how strong you need it. For a given armor thickness (or mass), diamond should give you the best per-area performance, but in my experience the falloff to the ~200W/m^2 materials is very slight (with the further falloff to ~20W/m^2 noticeable but still very small, particularly for radiators without much armor). Amorphous carbon seems a very good choice (particularly for hot reactor radiators, where there are not a lot of other options), although I somewhat wonder if it is more prone to snapping under kinetic fire than less brittle materials. I found that silica aerogel is sufficient in most cases, but silver + diamond as a thin outer layer can improve performance even further. My 0.5mm silver/1cm diamond/2cm silica aerogel missile lasts longer than a pure 10cm aerogel missile Do note that your reflective armor weighs 4.75 times as much as the 10cm aerogel... that sort of substitution is useful if you need to save space (in which case you should also be testing polytetrafluoroethylene, or aramid fiber if you care less about cost), but if you are mass limited (as the rocket equation means you tend to be) aerogel is really hard to beat for small objects.
|
|
|
Post by blothorn on Oct 8, 2016 8:15:09 GMT
Given the rise of superlasers, I ran some formal tests on drone laser turrets against a decently high-power laser of my own (500MW violet laser putting out 19.2MW with a 3.7m aperture; 2290MW/m^2 at 240km). The drone's body armor was 1mm boron so that it went quickly after the turret went down. All engagements started at 250km, with a closing speed of 500m/s. Reference areal density was 11,000 kg/m^2 (that of my previously standard drone turret armor). The target turret was a 68cm inner radius laser turret (my Solar Flare drone). Armor | Time (5 drones) | Comments | 5cm polytetraflouroethelyne | 81s | | 5cm silica aerogel | 105s | (1/22 reference density) | 10cm silica aerogel | 221s | (1/11 area density) | 5.2cm amorphous carbon | 1s | Ouch. | 4.1cm aluminum | 0s | Results inaccurate because I cannot double-tap my stopwatch fast enough. | 2.6cm alpha-2 titanium aluminide | 9s | Best specific heat of anything with a decent melting point. | 4.4cm s-glass composite | 24s | | 4.2cm silicon dioxide | 22s | The quest for a cheap hybrid armor continues to fail. | 3.9cm basalt fiber | 139s | | 7.8cm aramid fiber | 375s | Quite impressive vs. the basalt. |
Conclusions: superlasers certainly do not confer immunity from drones (at least by turret sniping; radiator sniping is its own can of worms, which I may visit later.) Silica aerogel really reigns supreme; the other compound of note is aramid fiber (best combination kinetic+thermal, and best performance by thickness). I included aluminum (and titanium aluminide) because people keep pushing for reflective materials its heat capacity suggests it is better than silver (and copper/gold do not reflect violet light well); that strategy failed miserably, but is likely to do better on larger scales. (This test proved to favor low conductivity far more than the missile test.) Also note that while I was running this setup, I pushed about half of a flight of 20 150kg nukes with 1cm aerogel armor through the laser and obliterated the test ship. Ridiculous lasers may be able to beat laserless capital ships by turret sniping, but I am really not convinced that they change anything about the seeming superiority of ordnance. Also, before I switched to the 1mm boron armor I was using my standard 1cm silica aerogel/1cm boron; the laser failed to destroy 5 drones in 6 minutes, by which time the drones had entered easy railgun range (100km) at the leisurely pace of 500m/s. I would strongly suggest against not armoring your turrets because you think you can kill the enemy's lasers first; six minutes is plenty of time for a squadron of 2.5MW laser drones to work through 3cm of boron.
|
|
|
Post by blothorn on Oct 7, 2016 20:58:20 GMT
You can also get accidentally good results in very specific circumstances if you perfectly tune your velocity and armor so that enemy lasers detonate flak at a reasonable range.
|
|
|
Post by blothorn on Oct 7, 2016 19:09:42 GMT
It messes up some other pieces of logic, such as weapons with "targets shots" disabled (so as to not waste ammo on small targets) not firing on unmanned cap ships. This could be fixed--I want more flexibility with those options to, for example, make anti-drone missiles that cannot be spoofed by micromissiles.
More generally, there are maintenance concerns and such with unmanned ships that can be handwaved away for small, disposable drones that are harder to do for capships--one of the game's assumptions does seem to be that not everything can be automated.
|
|
|
Lasers
Oct 7, 2016 16:36:09 GMT
Post by blothorn on Oct 7, 2016 16:36:09 GMT
Of curiousity, are people actually using these near-zero-armor turrets in combat? I would expect them to be extremely vulnerable to countersniping.
|
|