artforz

You're not going crazy, I can reproduce it here. Take the fresh built version, stick the generator *below* the reactor. Magically works. Alternatively, add another 4k battery bank. Also works. What's going on here is different part OnStart/OnFixedUpdate order (added a bunch of debug prints to tell WTF is going on): fresh built: [LOG 14:49:22.778] Reactor OnStart [LOG 14:49:22.780] [ORS] Creating Resource Manager for Vessel FRTB fresh built version (ThermalPower) [LOG 14:49:22.780] [ORS] Creating Resource Manager for Vessel FRTB fresh built version (WasteHeat) [LOG 14:49:22.780] [ORS] Creating Resource Manager for Vessel FRTB fresh built version (ChargedParticles) [LOG 14:49:22.782] Warning: Fusion Reactor MJ deprived, consuming 3693.79996626526 ElectricCharge [LOG 14:49:22.813] Generator OnStart [LOG 14:49:22.813] [ORS] Creating Resource Manager for Vessel FRTB fresh built version (Megajoules) [LOG 14:49:22.814] [KSP Interstellar] Configuring Generator [LOG 14:49:22.815] [KSP Interstellar] Radiation Module Loaded. [LOG 14:49:23.156] Reactor OnFixedUpdate [LOG 14:49:23.160] Warning: Fusion Reactor MJ deprived, consuming 3693.79996626526 ElectricCharge <- completely depletes ElectricCharge here [LOG 14:49:23.160] Warning: Fusion Reactor power deprived [LOG 14:49:23.161] Warning: Reactor resource deprived, shutting down [LOG 14:49:23.162] Generator OnFixedUpdate [LOG 14:49:23.330] Reactor OnFixedUpdate [LOG 14:49:23.331] Generator OnFixedUpdate ... stripped version: [LOG 14:52:31.396] [KSP Interstellar] Radiation Module Loaded. [LOG 14:52:31.425] Generator OnStart [LOG 14:52:31.425] [ORS] Creating Resource Manager for Vessel FRTB Stripped Version ktec (Megajoules) [LOG 14:52:31.425] [ORS] Creating Resource Manager for Vessel FRTB Stripped Version ktec (WasteHeat) [LOG 14:52:31.425] [KSP Interstellar] Configuring Generator [LOG 14:52:31.454] Reactor OnStart [LOG 14:52:31.454] [ORS] Creating Resource Manager for Vessel FRTB Stripped Version ktec (ThermalPower) [LOG 14:52:31.454] [ORS] Creating Resource Manager for Vessel FRTB Stripped Version ktec (ChargedParticles) [LOG 14:52:31.454] Warning: Fusion Reactor MJ deprived, consuming 3693.79996626526 ElectricCharge [LOG 14:52:31.666] Generator OnFixedUpdate [LOG 14:52:31.666] Reactor OnFixedUpdate [LOG 14:52:31.691] Generator OnFixedUpdate [LOG 14:52:31.691] Reactor OnFixedUpdate ...

Done. Quite a few values changed slightly, thrust now takes into account that KSPi uses g0=9.82m/s instead of 9.81, reactor mass and TWR now include the reactors internal fuel.

Don't think it'll work, from what I can tell FNGenerator would have to check part.srfAttachNode

Yes. It should be the same in the official 0.11. To be able to swap fuel in a reactor, it has to have less or equal to 0.01 actinides in it (see https://github.com/FractalUK/KSPInterstellar/blob/master/FNPlugin/FNNuclearReactor.cs#L71) It's also kinda sorta documented in the wiki: https://github.com/FractalUK/KSPInterstellar/wiki/Reactors#swap-fuel

Where? in onUpdate, thermal_power_received and total_power are actual energy produced this update, not the max (which would be ThermalPower * TimeWarp.fixedDeltaTime). in onStart, tritium_rate later gets multiplied by ongoing_consumption_rate. which is the persistent copy of total_power_ratio, set to total_power / ThermalPower / TimeWarp.fixedDeltaTime by onUpdate... edit: For the charged particle ratios... no clue. I *think* D/T is supposed to be about 0.199, so likely 0.20 is more correct than 0.21. For D-He3 it depends on a lot of factors, and I got no idea how/why Fractal picked that particular value. What's a lot more significant is the fact that with 0.11, running a fusion reactor in D-He3 or He3-He3 mode is rather annoying due to https://github.com/ArtForz/KSPInterstellar/commit/2cef62f679532982a732c0b7771edf5687075ec4 Oh, and the whole "Fusion reactors consume only their own T supply, while breeding spreads it evenly over all containers" thing also doesn't help making Tokamaks particularly user friendly... Nor does the fact that a Vista has no such qualms about where it gets its D/T from, it happily grabs it from every container on your vessel including reactors.

For the tritium breeding, I'm pretty sure it's a bug in the amount calculation. As it currently is is, we basically have trit_rate = <constant> * max total power * total active% when on rails and trit_rate = <constant> * current thermal power * (1-chargedparticleratio) when focused... put another way, if we fully load the reactor we get rails trit_rate = <constant> * max power vs focused trit_rate = <constant> * max power * (1-chargedparticleratio)2 That can't possibly be right. @Fractal_UK: Is breeding rate supposed to be based on total output power times 1-chargedParticleRatio? Then https://github.com/ArtForz/KSPInterstellar/commit/7034d39c79fe174ea25dadae1985f114dad2142d should be the proper calculation. If it's supposed to be based on total output power times 1-chargedParticleRatio when on rails and thermal output power when active, then the change in onStart is still correct and in OnFixedUpdate tritium_rate = (float)(thermal_power_received / TimeWarp.fixedDeltaTime / 1000.0f / GameConstants.tritiumBreedRate)*(1-chargedParticleRatio); should be tritium_rate = (float)(thermal_power_received / TimeWarp.fixedDeltaTime / 1000.0f / GameConstants.tritiumBreedRate);

Nice. Hmm... you got extraneous OpenResourceSystem_1_1_0.dll in GameData/WarpPlugin/Plugins ?

https://github.com/FractalUK/KSPInterstellar/blob/master/FNPlugin/FNGenerator.cs#L342-345 https://github.com/FractalUK/KSPInterstellar/blob/master/FNPlugin/FNNozzleController.cs#L432-435 I'm no mind reader, but that looks pretty deliberate to me. In versions < 0.10.2 the reactors themselves converted unused ChargedParticles to ThermalPower, but I think that caused issues with reactors not going to 100% in some situations.

Bug. In FNNuclearReactor.cs RefuelUranium()/RefuelThorium() double spare_capacity_for_uf6 = uf4.maxAmount - actinides.amount; should be double spare_capacity_for_uf6 = Math.Max(uf4.maxAmount - uf4.amount - actinides.amount, 0); similar for Thorium.

Those numbers look correct. Thermal Generators can use ThermalPower and ChargedParticles, and for determining their max power they assume they can get all of both. If there is a DC generator, that gets dibs on the ChargedParticles, leaving only ThermalPower for the thermal gen. So let's look at that upgraded Sethlans2: 1285MW total with 82% of it ChargedParticles. So that's 1053.7 ChargedPower and 231.3 ThermalPower. With just a KTEC, it gets all of both and converts it. Assuming enough radiators and sitting on the launchpad, it's ~55.7% efficient. 1285 * 0.557 = ~716. So that looks right. With just a DC, it gets the 1053.7 ChargedPower, converts that at 85% efficiency. 1053.7 * 0.85 = ~ 896. The 231.3 ThermalPower aren't used at all. So that looks right, too. With a KTEC + DC, the DC still gets the ChargedPower, producing ~896MW. The KTEC gets the ThermalPower, 231.3 * 0.557 = ~129MW. Again, looks right. Oh, and with KTEC + DC, the generators really can output over 1.6GW combined. For half a second or so until the ChargedPower/ThermalPower buffers are depleted. So I'm guessing the problem is "100% transmitter setting is the sum of peak output of all active generators"?

At its peak 3500K, a upgraded KSPI radiator is hotter than the filament in a halogen bulb. Those do actually... glow. Less snarky, let's go with the wiki figures for a standard fold-out: 3403.64MW, 400m^2, 3500K. At ~ 8.5MW/m^2 and assuming a emission spectrum close to a black body, that'd be a rather impressive source of yellowish-white light...

Confirmed, AM reactors produce way too much power when supply limited. The more oversized the reactor relative to the amount supplied by collectors, the more pronounced the effect. see https://github.com/FractalUK/KSPInterstellar/pull/37 for a tentative fix