|
Post by apophys on Oct 7, 2016 16:58:55 GMT
Thanks, I figured there was a design flaw somewhere, but I couldn't put my finger on it. Sadly, the problem with too wide inner turbopumps is the massive increase in mass, so a nice balance needs to be kept. My goal here is max power per unit of mass. As my current design only runs at 11.6% efficiency there's still a bit of work to do. I didn't realize that gigantic 12.4 GW reactor was meant to be practical. Well then. Since there's actually a desire for even higher power than my 1GW reactor, I feel compelled to upscale my design. Here's 10 GW. It comes at 151 t of mass, so it's marginally more mass efficient than my 1GW (at the cost of ballooned volume, so armoring will suck). In the course of building this, I find that a slightly lower efficiency than the maximum slightly improves the stats here, because of the absurd draw of the inner turbopump at these scales. The thermocouple can be rearranged to have high radius and low height instead of the current high height and low radius. ThermoelectricFissionReactorModule 10.1 GW Thermoelectric Fission Reactor ReactorCoreDimensions_m 0.25 0.1 NuclearReactor Coolant Sodium Moderator Diamond ModeratorMass_kg 1 Fuel U-235 Dioxide FuelMass_kg 110 FuelEnrichment_Percent 0.97 ControlRodComposition U-233 Dioxide ControlRodMass_kg 99 NeutronReflector Diamond ReflectorThickness_m 0.6 AverageNeutronFlux__m2_s 2e+020 InnerTurbopump Composition Amorphous Carbon PumpRadius_m 2.3 RotationalSpeed_RPM 480 ThermocoupleInnerDimensions_m 13 30 Thermocouple PTypeComposition Tungsten NTypeComposition Tantalum Length_m 0.001 ThermocoupleExitTemperature_K 2500 OuterCoolant Sodium OuterTurbopump Composition Boron PumpRadius_m 0.75 RotationalSpeed_RPM 800
|
|
|
Post by apophys on Oct 7, 2016 11:45:15 GMT
What are people's opinions regarding cooling via the inner pump vs the outer pump? How does one have to think about it? If the inner pump is done badly, the fuel will melt at lower core temperatures than it otherwise would (because the heat isn't spread effectively). As an example of this effect, see the 12.4 GW reactor posted by tukuro. You want a high core temperature for optimal power generation. This makes the inner pump much more important, and you can't shave off quite as much mass from it as from the outer one when optimizing. Keeping the same core temperature, the outer pump can be somewhat shrunken and driven at slightly higher speeds, while the thermocouple is increased. This reduces overall mass (mostly sodium) while the dimensions and power remain about the same. Also, get a forum account, you won't regret it. I'm screwing around with making decoy drones, essentially just reactors, heatsinks, and a high accell rocket with enough delta V to dodge missiles for a little while, does anyone have any small light reactors that put out a LOT of heat? Heat generation depends only on the fuel and neutron flux. Unfortunately, this heat still needs to be dumped out the hard way, so there's no shortcut. The best you can do is run your reactor at a higher output temperature to shave some radiators. How much heat are you thinking about? Another thing you can try is not having any acceleration at all, and just launching reactor+radiator from coilguns. The heat source will be closer and thus more effective as a decoy.
|
|
|
Post by apophys on Oct 7, 2016 10:25:15 GMT
This was the largest 2500k I was able to design. Has anyone managed to get more? More you say? How about 37.6 GW? The limiter is turbopump energy consumption. If you spin them faster to cool more, they eat more power than the cooling helps with.
|
|
|
Post by apophys on Oct 7, 2016 9:41:01 GMT
So there has to be a balance, but I will experiment with 2500K reactors. Are there any downsides to them? The higher the output temperature of a reactor, the less raw power you're able to get out of it, because the temperature difference from the core (~3100) to radiator lessens, and that increases reactor mass versus power generated. This is made completely irrelevant by the fact that (with good reactor designs) radiators significantly outweigh reactors, and radiators get better at their job at higher temperatures. Higher output temperatures reduce the heat:power efficiency of a reactor, for the same reason. I.e. you're producing more heat for a given amount of power. This can make heat decoys rather impractical. A higher radiator temperature brings it closer to the melting point, making it easier to break with nukes and lasers. This is not a big deal though, because higher radiator temperatures also make them many times more efficient at radiating heat, so you can afford to place multiple redundant ones. Also, materials like amorphous carbon or diamond can be chosen, which have very high melting points. Overall, it's totally worth it.
|
|
|
Post by apophys on Oct 7, 2016 8:35:38 GMT
Certainly more than 3 people here play FTD. For anyone who doesn't know, that's From the Depths. Although for me, CoaDE is taking huge priority right now.
|
|
|
Post by apophys on Oct 7, 2016 6:58:36 GMT
I think turrets can be modules to themselves, rather like how we currently have launcher modules for missiles. You pick a 'payload', in the case of the turret it would be a laser, coil gun, rail gun or chem gun. Then you pick actuator type: electro-mechanical, pneumatic, or hydraulic. Electro-mechanical systems are the simplest: an electric motor drives a gear attached to the turret along your desired axis. It's lightweight, but it rapidly loses effectiveness as the mass of the payload increases and the torques needed get out of hand. Pneumatic systems use a pneumatically driven rack and pinion to drive the turret gear. They can generate more torque per input power, but they need a compressor - which means you use either an electric motor to drive it, or you have to divert some of your rocket's exhaust gasses. Hydraulic is just a beefier pneumatic system, that needs more power and can drive the big anti-ship turrets. I'm not an expert on any of these systems, so maybe you don't need all three, but I thought I'd put it out there. Finally there's an armoring section, identical to the ship one where you can design armor layers to cover it all up. I think pneumatic and hydraulic can be combined, and we can pick the working fluid from a short list. Add reaction wheels as an actuator option; this system should replace the current forced wheels, but they should remain a possibility. Payloads should be diversified to also include any thrusters. We can deal with the structural effects of recoil and other forces with one option for part thickness. The same option would alternately allow us to customize our reaction wheels if we choose those (their radius can be free to snap to the maximum possible, I just want to be able to disconnect wheel mass from turret size). I also would like an option to stick it out from the hull by X distance, for the nice battleship look (thus being able to point all guns on the entire ship directly forward). This would also increase our available angle of fire downward toward the hull, so we can have more than a hemisphere of coverage if we desire. If sticking it out really far, the connecting bit should be the radius of the turret ring (which we choose).
|
|
|
Post by apophys on Oct 7, 2016 5:44:54 GMT
By the way, here's the current version of my 1GW superlaser. Intensity at 240km is 1240 MW/m 2, mass is 46.9 t, and cost is 514 kc. Armor is 3 cm of boron. LaserModule 1.000 GW Nd:YAG Green Laser ArcLamp GasComposition Krypton EnvelopeComposition Fused Quartz PowerSupplied_W 1e+009 Radius_m 0.01 CavityWallComposition Silver CavityCoolantComposition Hydrogen CavitySemimajorAxis_m 1 CavitySemiminorAxis_m 0.99 GainMedium Nd:YAG OpticalNodes 7000000 LasingRodRadius_m 0.014 Mirror Composition Silver OutputCoupler Composition Fused Quartz CoolantTurbopump Composition Boron PumpRadius_m 0.9 RotationalSpeed_RPM 460 CoolantInletTemperature_K 1200 FrequencyDoubler NonlinearOptic Composition Silver Gallium Selenide OpticLength_m 0.026 OpticRadius_m 0.025 ApertureRadius_m 2.4 FocusingMirror Composition Silver Turret InnerRadius_m 5.1 ArmorComposition Boron ArmorThickness_m 0.03 ReactionWheels Composition Lead RotationalSpeed_RPM 72 EngagementRange_km 250 TargetsShips true TargetsShots true
|
|
|
Post by apophys on Oct 7, 2016 4:06:19 GMT
Try this laser. More efficient, smaller, lighter, and enormously increased intensity (123 MW/m 2 at 200km); although with less angle (45 deg.) and traverse speed (45 deg./s). Edit: an immediate update, now weighs even less at 10.6 tons. LaserModule 100 MW Titanium:Sapphire Violet Laser 2 ArcLamp GasComposition Xenon EnvelopeComposition Fused Quartz PowerSupplied_W 1e+008 Radius_m 0.06 CavityWallComposition Silver CavityCoolantComposition Hydrogen CavitySemimajorAxis_m 1 CavitySemiminorAxis_m 0.99 GainMedium Titanium:Sapphire OpticalNodes 9000000 LasingRodRadius_m 0.079 Mirror Composition Silver OutputCoupler Composition Fused Quartz CoolantTurbopump Composition Boron PumpRadius_m 0.9 RotationalSpeed_RPM 46 CoolantInletTemperature_K 1200 FrequencyDoubler NonlinearOptic Composition Silver Gallium Selenide OpticLength_m 0.026 OpticRadius_m 0.0089 ApertureRadius_m 1.6 FocusingMirror Composition Aluminum Turret InnerRadius_m 3.4 ArmorComposition UHMWPE ArmorThickness_m 0.02 ReactionWheels Composition Cadmium RotationalSpeed_RPM 44 EngagementRange_km 210 TargetsShips true TargetsShots true
|
|
|
Post by apophys on Oct 6, 2016 8:09:42 GMT
Some notes from me: - Irradiance can be increased greatly by buffing aperture (and turret size); output power has no simple increase once you hit 1GW power in. So I care somewhat more about output power for my lasing materials. - In my testing, Nd:YAG+krypton consistently beats Ti:Sapphire+xenon and Nd:GGG+krypton in terms of output power, which in turn consistently beat everything else. Nd:YLF+krypton exists, but the melting point makes me entirely disregard it, as it has no outstanding redeeming qualities. - I can get Ti:Sapphire to pleasant efficiency levels beyond what Nd:YAG can accomplish, but I have so far not been able to do it without a large cavity, costing enormous amounts of hydrogen mass. - I see no reason to use anything other than hydrogen as coolant. - I suspect that there's an optimal irradiance. If you're able to burn through a weapons module, you don't need to make targeting harder with a smaller focal point. It would be hilarious if you could cut ships in half with an extremely focused beam like a sword, or core drill them, but I suspect it doesn't work that way, and you'd be better off barely burning away the modules and whipple shield.
The big one:
- For me, using molybdenum instead of silver for the shell, and diamond instead of fused quartz for transparent bits, cuts output power by ~55%. It saves ~90% of the radiator surface area with a 2240K output temp. Looks like it's worth it, right?
Well, with reactors at 2500K, high-temperature laser radiators take about 22% of the area of the reactor radiators used to provide the power the laser consumes. Switching to a higher laser operating temperature saves you 62% of your original total radiator area. Considering the 55% loss of power, it's basically scaled down without any savings, and the mass of the laser+reactor is not scaled. So this is actually slightly worse.
I hope I haven't made an error in my arithmetic, because this is a big deal.
A higher reactor temperature would change this of course, but not by much, I think.
|
|
|
Post by apophys on Oct 6, 2016 5:37:57 GMT
Wow, that's really tiny. I'm avoiding building spammy things of that scale because of framerate reasons. I think not having a reactor, or a battery, on a missile is unrealistic. Gimbals in every module other than NTRs and combustion rockets require electricity. I'm expecting that the free gimballing is either a bug hiding in plain sight or temporary until batteries exist. It's not much electricity they should normally require, but nevertheless, it is nonzero. Right now there's no reason not to crank up the spin speed as high as the material can survive it, and there's no reason to use heavier materials for energy efficiency.
|
|
|
Post by apophys on Oct 6, 2016 3:20:52 GMT
Unfortunately, decane/LOX has a lower limit on propellant mass of about 45.0 kg due to the stoich ratio and the 10.0kg minimum tank size. I'll probably try a monoprop design tonight. I really want to keep the small diameter, but I've noticed that puts a pretty hard cap on your gimbal angle, which has some pretty significant repercussions in terms of maneuverability. I'd love to get my missiles and launchers down to the size where drones can use them, though. Try decane resistojets. Easily scalable downwards as small as you need, and over 6.3 km/s exhaust velocity if you build well. Of course you'll need a nice reactor, but with current capabilities, that's not a problem. Something to get you started. 1 MW input, 6.35 km/s exhaust, 26 kN thrust & 1.97 kg (1340 TWR). Cost 15c. ResistojetModule 6.35 km/s 1.00 MW Decane Gimballed Resistojet PowerSupplied_W 1e+006 Propellant Decane CoilComposition Tantalum Hafnium Carbide ChamberLength_m 0.1 CoilRadius_m 0.0001 ThermalRocket ChamberComposition Diamond ThroatRadius_m 0.01 ChamberWallThickness_m 0.00018 ChamberContractionRatio 6 NozzleExpansionRatio 90 NozzleExpansionAngle_degrees 7 RegenerativeCooling_Percent 1 Injector Composition Lithium PumpRadius_m 0.095 RotationalSpeed_RPM 80 Gimbal InnerRadius_m 0.14 ArmorComposition Lithium ArmorThickness_m 0.0001 ReactionWheels Composition Lithium RotationalSpeed_RPM 2200 GimbalAngle_degrees 18
I'm in the process of scaling it up for capital ship thrust, since power is so damn cheap.
|
|
|
Lasers
Oct 6, 2016 2:32:37 GMT
Post by apophys on Oct 6, 2016 2:32:37 GMT
I'm trying to see if I can get a Near Ultraviolet laser going - the advantage is that, for a give aperture, and output power, it seems to have a higher intensity at any given range. The catch is that it's less efficient overall. I don't really understand how the arc lamp spectra or pumping efficiency is calculated, so.... I'm not an expert in the field, but this is my take. The vertical red lines are the resonant frequencies of the gain medium. You want the green spectrum graph of the gas (the frequencies it naturally emits when blasted with electricity) to be high where at least one of those red lines intersect it. Ti:Sapphire matches well with xenon because the second red line hits a high point on the spectrum, and the first one is okay too. Krypton emits frequencies in a range that several materials match up with. Nd:YAG gets multiple nice intersections. Putting more power into a laser makes gas emit more high-frequency radiation, and since all the materials we have resonate in the very low end (except ruby), it makes the laser marginally less efficient. You can see the green graph sway. A larger radius of the gas tube makes pumping more efficient for some reason (but this worsens cavity shape, which you can otherwise compensate for). I find the minimum radius to be unequivocally best for Nd:YAG + krypton. Haven't tested others yet, because this combo makes a nice compact cavity with fine efficiency. None of this has anything to do with intensity at range. Higher frequency light simply carries more power, that's where the intensity is. But our high frequency options are currently quite poor.
|
|
|
Lasers
Oct 6, 2016 1:31:45 GMT
Post by apophys on Oct 6, 2016 1:31:45 GMT
To scale up a laser in power, just give it more power, and tweak the frequency doubler to get 100% again (and not melt). If anything else is melting, pump more coolant. Ah, you also will probably need to increase the Optical Nodes to bring M 2 back down to 3.00 Of course you can do the reverse to scale down any laser in power, like my 1GW to a 320MW.
|
|
|
Lasers
Oct 6, 2016 1:02:26 GMT
Post by apophys on Oct 6, 2016 1:02:26 GMT
As near as I can tell, what's happening is that the lens aperture and the reaction wheel for the turret are competing for space; if you increase the size of the aperture you will decrease the amount of mass in the reaction wheel. If the 8m aperture is a tighter 'fit' for the turret than the 7m aperture is for the smallest turret that will fit it, the 7m one will be heavier even if it's smaller. You'll also lose turret speed, but you don't need a lot of turret speed to engage targets that are 200 km away. But this highlights what a big design problem the turrets themselves are, even independently from the laser. In any case, my general finding would be that you want to make turrets in the range of 4m across, but that this is heavily dependent on the turret having the right "fit". Yeah, that's the effect I was exploiting to make my ultralight 10.3m turret. It is an extremely close fit, so much so that I'm using iridium reaction wheels to get a decent power usage from them with the traverse speed I desire (45 degrees per second). The aperture for this point is 2.4 m. I noticed the effect when playing with gimballing thrusters earlier, in those the injector is competing for space with wheels.
|
|
|
Post by apophys on Oct 6, 2016 0:34:31 GMT
Ooh. Having zero armor certainly makes things easier... Nice find.
Hm... 23m aperture is the largest that fits on a turret, because turrets go up to 50m only... but apertures go to 100m... So can we do away with the turret entirely by undersizing it? Aim with the ship?
If so, I could hit 946,000 MW/m2 at 240 km for 1.07 kt and 26 Mc ... But wait a minute, it's only outputting 19.6 MW. That's 20 square millimeters. How much damage would that even do...
|
|