FreeThinker

[1.7.1/1.6.1/1.5.1/1.4.5/1.3.1] KSP Interstellar Extended 1.21.9 Continued Development Thread

Recommended Posts

Never used them personally. It was one of the last things Fractal_UK implemented before he went away, and I'm not quite sure exactly how he made them to work.

A *REALISTIC* implementation would basically use ChargedParticles as the only propellant. That is, it would (SLOWLY) convert Uranium into propulsion, at an ISP of over 50,000 seconds! However, the Thrust for this would be DEATHLY low- as in, on the order of magnitude of Solar Sails. So, I'm not quite sure how Fractal_UK *ACTUALLY* made these to work...

It only runs off of charged particles because that's the only form of power you can use to heat your propellant directly. The rest of the output comes in the form of neutrons which heat the solid portions of the reactor, thus the temperature you can heat your propellant to is limited by their failure point.

Magnetic nozzles have always been a little strange in this mod, ever since their implementation. I think, in their most ideal sense, they're the answer to everyone asking about fission fragment rockets (which, at the expense of producing practically no thrust at all, could conceivably have Isps as high as 1.5 million seconds). In practice, though, they'd suffer the same exact problems as other realistically-low-thrust engines in KSP if they were proper fission fragment rockets: no one wants to sit around and wait for a 30-newton thruster to change the orbit on their 50-ton ship.

Solar sails have partly addressed this problem by allowing your ship to experience forces while on rails (during timewarp), but there are problems with doing this for regular engines. Fractal seems to have done it by using liquidfuel to 'gear down' the engine - adding extra mass to the reaction, so the exhaust velocity drops and the thrust increases. The exact mechanics of how this works for the charged particles released by various reactions (ie fission vs. fusion) are a little dubious, but it's ostensibly valid physics, and suffers only from the fact that the <1 charged particle ratio of reactors means that the thrust power when using these engines is limited. I think they might see more usefulness if the 'gearing' were variable like on the DT Vista - keep the power consumption constant, and make lower throttle equal higher Isp and lower thrust.

The other main problem with magnetic nozzles as implemented in KSPI, of course, is that they don't work on antimatter reactors. Ever since the thermal nozzle nerf (in which Fractal apparently decided that thermal nozzles are thermal nozzles only, and we can't pretend they're plasma-core rockets or other such things), antimatter propulsion has been all but useless- there's no way to get a decent Isp out of it, and thanks to some recent bugs, you can't even really get decent thrust out of it either (most of the power just vanishes into thin air). Antimatter reactors seem like they would be the best niche for magnetic nozzles - the huge power output means you can enjoy respectable thrust along with your very high Isp - but, for whatever reason, Fractal decided that antimatter won't work with magnetic nozzles, because apparently antimatter reactions don't produce charged particles that can be deflected with a magnetic field. Indeed, the antimatter reactors in the game don't produce charged particles, so on some level, it makes sense.

Except it doesn't, because antimatter reactions do produce charged particles. The charged pions emitted from the reaction are short-lived and quickly decay into gamma rays, but not so quickly that redirecting them with powerful magnetic fields and gaining incredibly powerful propulsion at pretty good efficiencies isn't possible.

Of course, if we're talking about direct-annihilation beam-core antimatter propulsion, we'll run into two main problems: one, the thrust will still be absolutely microscopic for the power levels in question ((405 GW / 100 Mm/s * 2)*60% eff =~ 1.2 kN, which is higher than a fission-fragment but still not a lot when you're pushing a giant reactor around), and two, you'll need prohibitive quantities of antimatter and no meaningful quantities of anything else, since your propellant is equal parts matter and antimatter - that's going to kill your mass ratio. Plus, your Isp is going to be so ludicrous as to be unbalanced. 10 million seconds? KSP doesn't have interstellar flight just yet!

The obvious answer to both of these issues is the plasma-core antimatter engine. Just add more hydrogen - your thrust goes up, your Isp goes down, and you've got a useful and more or less game-balanced engine. Oh hey - adding hydrogen to fix the low thrust issue. Sound familiar? :D

Ultimately, I'd be inclined to suggest that antimatter reactors be given a charged-particle ratio. It would make them work easily and realistically with magnetic nozzles, which would bring antimatter propulsion back to its former glory in this mod, and it would also be pretty true-to-life in how antimatter reactions would work.

Edited by GreeningGalaxy

Share this post


Link to post
Share on other sites
On jets you missed (3), which is that the jets don't care about current atmospheric density, whereas in reality thrust is proportional to both velocity and to density (in fact, thrust is closely proportional to Q, although taking compressibility into account).

The effects of atmospheric density are only due to how that effects the internal airspeed and temperature within the jet engine. With a precooler (a part that is available in KSP-Interstellar) these effects become *much* less significant up until you reach much higher compression factors...

Viewed from a scientific perspective, there is no difference between a jet engine and a rocket (which is actually technically a sub-class of jet engine) except the internal airspeeds and temperatures. There is nothing magical about flying at high altitude that automatically makes a jet engine less productive- if you cool the intake airflow to the same speeds+temepratures as at lower altitude, the performance will actually be *better* (due to reduced atmospheric compression of the exhaust-stream, although a very minor effect with the high Exhaust Pressure of most jet engines...)

Regards,

Northstar

Just taking a look at your screenshots, I think everything looks about right as-is.

For the second picture, with the engine producing 136 kN at 195.5 s, sticking that into the thrust power equation gives a total thrust power of about 130 megawatts:

(136,000 N * 195.5 s * 9.81 m/s/s) / 2 =~ 130,000,000 W = 130 MW

Given that the fully-upgraded dusty-plasma reactor of that size puts out a max power of 142 megawatts, a 130-megawatt thrust power seems to suggest rather generous values for conversion efficiency,

Your numbers are way off for the power of reactors. The reactor that Thermal Turbojet is attached to produces over 800 MW of ThermalPower. Are you running the latest version of KSP-Interstellar Extension Config? The fission reactors received a MAJOR buff to their ThermalPower production to bring them in line with real-world values for reactors of that size/volume (based on data from Timberwind, SNTP, NERVA, etc.)

It only runs off of charged particles because that's the only form of power you can use to heat your propellant directly. The rest of the output comes in the form of neutrons which heat the solid portions of the reactor, thus the temperature you can heat your propellant to is limited by their failure point.

Fission Fragment Rockets don't heat propellant- they expel the ChargedParticles (which are at temperatures of millions of degrees before hitting any other atoms/molecules) directly.

Regards,

Northstar

- - - Updated - - -

the thrust power equation

Also, the equation you're referring to is:

Power = 1/2 * Thrust * Exhaust Velocity

it is originally derived from the following equations:

E = 1/2 m * v2

Thrust = Mass Flow Rate * Exhaust Velocity

Notice the division by 2? Your calculations are all off because you forget to divide by 2. The Thermal Turbojet (connected to an antimatter reactor, I assume, based off the ISP?) is not getting more than 100% conversion efficiency. The ThermalPower of the reactor would back-calculate to 3 GW based on the proper implementation of the equation (in which you must divide by 2).

Regards,

Northstar

I don't think I have any reason to believe that it would be linear - I haven't tested un-upgraded reactors, larger reactors, or different methods of power generation (antimatter, AIM, fission, etc), nor more/less intakes. I think what would be ideal is to find the formulas that govern thrust and Isp on thermal engines and change them to make them behave realistically to begin with, but as I've never worked with this kind of thing before, I have no idea how hard that would be. In the short run, some kind of adjustment to just make sure the thrust power of the engine doesn't go above the power of the reactor attached to it would be better than nothing, even if it acted as a hard limit.

I'll do some more tests tomorrow and see if I can figure anything else out.

GreeningGalaxy, as I point out in my other post, your math is incorrect. You forgot to divide by a factor of two- so you overestimate the Thrust Power of the engines you look at.

Regarding the behavior of thermal engines, they are governed by a simple conversion of thermal energy into Thrust via "E = 1/2 m v2", which is the equation that has been at the base of our mass all along. After correcting for some inefficiency in the conversion, you get the performance we currently have for our thermal engines.

I respectfully suggest you look at how I went about calculating the expected behavior of the thermal engines (it is well documented earlier in this thread and in the KSP-I 0.90 port maintenance thread) instead of assuming I made a mistake in my calculations.

Regards,

Northstar

Edited by sal_vager

Share this post


Link to post
Share on other sites

Notice the division by 2? Your calculations are all off because you forget to divide by 2. The Thermal Turbojet (connected to an antimatter reactor, I assume, based off the ISP?) is not getting more than 100% conversion efficiency. The ThermalPower of the reactor would back-calculate to 3 GW based on the proper implementation of the equation (in which you must divide by 2).

You'll find that she actually didn't. If you read again;

0.5 * 1500 kN * 820 s * 9.81 m/s/s =~ 6,000,000,000 W = 6 GW (/ 5.5 GW = 110% conversion efficiency.)

The maths is correct. Run the numbers yourself: 1500 kN, at 820 seconds, is 6 GW.

Edited by Iskierka

Share this post


Link to post
Share on other sites
Fractal decided that antimatter won't work with magnetic nozzles, because apparently antimatter reactions don't produce charged particles that can be deflected with a magnetic field. Indeed, the antimatter reactors in the game don't produce charged particles, so on some level, it makes sense.

Except it doesn't, because antimatter reactions do produce charged particles. The charged pions emitted from the reaction are short-lived and quickly decay into gamma rays, but not so quickly that redirecting them with powerful magnetic fields and gaining incredibly powerful propulsion at pretty good efficiencies isn't possible.

Well I want to change that. The only thing I need is some numbers and some justification

- - - Updated - - -

The obvious answer to both of these issues is the plasma-core antimatter engine. Just add more hydrogen - your thrust goes up, your Isp goes down, and you've got a useful and more or less game-balanced engine. Oh hey - adding hydrogen to fix the low thrust issue. Sound familiar? :D

Basically this is our Magnetic nozzle, except with more Isp and trust.

Edited by FreeThinker

Share this post


Link to post
Share on other sites

Regarding the gameplay issue of low (as in realistically low not KSP ion drive low) thrust engines, it can be partly remedied with Orbit Manipulator , where the limit is shifted from timewarp to orientation adjustment.

Share this post


Link to post
Share on other sites
GreeningGalaxy, as I point out in my other post, your math is incorrect. You forgot to divide by a factor of two- so you overestimate the Thrust Power of the engines you look at.

Check the math again - I didn't, in fact, miss the division by 2 in any of the calculations I've done so far.

Your numbers are way off for the power of reactors. The reactor that Thermal Turbojet is attached to produces over 800 MW of ThermalPower. Are you running the latest version of KSP-Interstellar Extension Config? The fission reactors received a MAJOR buff to their ThermalPower production to bring them in line with real-world values for reactors of that size/volume (based on data from Timberwind, SNTP, NERVA, etc.)

I have installed the updated version (just now, as it happens). You're right, the power output I was giving for the particle bed reactor was pretty low, but, based on my own tests, I still don't believe that the engines are remotely under-performing. I've managed to redo my tests of the 2.5m thermoturbojets and still find them to be producing too much power at peak (I got exactly the same numbers as before). I also did tests of the 1.25m reactors, and everything I tested (particle bed, dusty plasma, and fusion) was producing almost equal thrust power to reactor power at peak performance. They probably need a significant nerf to account for the fact that you aren't going to get more than about 80% total efficiency out of that kind of engine, but they're at least not pulling power out of thin air. I would definitely still say that the thermal jets are in no way in need of a buff, though - they now have quite significant TWRs with small engines and are pushing the limits of theoretical efficiency, so increasing their power would benefit neither realism nor game balance.

Well I want to change that. The only thing I need is some numbers and some justification

Justification? Well, collecting antimatter in the mod's current state is pretty much pointless, because there's no nice powerful propulsion system to use with it ever since Fractal nerfed thermal rockets. Just for gameplay reasons, a high-thrust high-ISP antimatter propulsion system would be great to have.

As for numbers, I can probably help a little more with that.

Project Rho's page on Exotic Weapons has a bit about the kinds of things antimatter gets up to during annihilation. Based on what I found there, it looks like approximately 55% of the energy from an antimatter reaction will show up as charged pions, while the rest is going to end up as gamma rays (which, for the sake of simplicity, I think the mod refers to as thermal power). Of course, there are other small details to consider as well - mainly, you're not going to thermalize all the gamma rays. What that means is that, say, the 405-gigawatt antimatter reactor, in order to produce 405 gigawatts of extractable power, might have to annihilate considerably more than 405 gigawatts' worth of antimatter. The gammas, particularly those formed as a result of neutral pion decay, probably are going to be tricky to trap. Project Rho seems to be of the impression that about 1/3 of the energy of the reaction will be from neutral pions, so if we assume that half the neutral pions and their subsequent decay-product gamma rays escape the reactor without providing useful heat, that means that only ~66% of the antimatter energy is captured. Furthermore, if we assume that all of the charged pions are captured, that means that our charged particle ratio relative to the total power is going to be something like 0.55/0.66, or about 83%. Assuming that from that point on, the charged pions are expelled from the magnetic nozzle with near-perfect efficiency (an assumption that the mod apparently does make for the other kinds of charged particles), then we'll see quite handsome engine performance (in the neighborhood of 320 gigawatts, if you make the same assumptions I did.)

As of now, antimatter reactor consumption is a little strange. If we assume that the speed of light c is the same in KSP as it is for our universe, the 15 milligrams/second draw of the 405 GW reactor comes out to E=2*0.000015*299792458^2 or about 2.7 terawatts. That's a rather low efficiency, about 15%.

However, that's forgetting one small detail: For the Alcubierre drives contained in the mod, the speed of light for the itty-bitty kerbal universe is assumed to be 1/10 of the one in our universe. Originally, this was just to keep the drives from being unreasonably fast relative to the system, but it also has the implication of reduced energy/mass equivalence. Using c*0.1 for our antimatter yield gives us E=2*0.000015*29979245.8^2, or only about 27 gigawatts! In short, that antimatter reactor is getting a lot of power out of nowhere if you assume that the speed of light is consistent between the antimatter reactors and the Alcubierre drives.

To find the antimatter draw that would really be required for a 405 gigawatt reactor assuming the above 66% total efficiency, we could use M = E/2*c^2, so M = (405 GW/0.66 eff) / (2*29979245.8^2) = 0.000341 kg/s or 341 milligrams per second. It's a little high, but I think it's reasonable - if it's really an issue of game balance, the capacity of the antimatter containers could always be bumped up a bit.

All this physics is making me hungry.

--Addendum--

I just remembered - the differing value for c would also drastically affect the yield of fission and fusion reactions, because the energy released is due to the change in mass defect of the nuclei. Since it would probably wreck everything to try and fix all the reactors to account for this, it might make more sense to keep antimatter yield calculations using normal real-world c (in which case the antimatter draw to produce 405 gigawatts at 66% efficiency would drop to 0.0000034 kg/s or 3.4 milligrams per second). You're the modder, it's your call. :P

Edited by GreeningGalaxy

Share this post


Link to post
Share on other sites

Was able to *partially* reproduce your results GreeningGalaxy:

vl4n7Ab.jpg

HNPom1r.jpg

145.4 kN of Thrust at 191.7 seconds ISP and 138 MW of ThermalPower!

It looks like there is an issue with how Thrust is being calculated with the Thermal Turbojets with reactors that can very their ThermalPower based on WasteHeat accumulation (namely the Particle Bed and Dusty Plasma reactors). With WasteHeat= 0 these reactors still produce the same excessively-low Thrust/MW levels I was observing before...

Or rather, the issue seems to be that these reactors are not actually dropping their ThermalPower production when WasteHeat accumulates to levels as low as displayed... The Thermal Turbojet I tested above got roughly the same Thrust as my plane from before, which was operating at 0 WasteHeat and thus had over 849 MW of ThermalPower available:

b4vHGMk.jpg

145.4 kN of Thrust at 190-ish seconds is reasonable for a 850 MW reactor (in fact, it's too low- hence why I stated a 47% increase to the Thrust modifier for Thermal Turbojets is in order), but is *NOT* reasonable for a 138 MW reactor.

FreeThinker, it seems we have a bit of a bug with the displayed ThermalPower levels not actually matching what is going on under-the-hood with some reactors. This needs to be sorted out before we can get any reliable data on what is actually going on with the Thermal Turbojets...

OK, one more thing.

GreeningGalaxy, I examined your one screenshot a little more closely, trying to figure out why your Thrust/MW is so far out-of-line with anything I'm getting, and I think I might have an idea (although, once again, you really, really, really need to post a lot more screenshots to confirm there is even an issue. At the very least, you should show the reactor's context menu in one screenie and the Thrust/ISP of the ThermalTurbojet in the very next one...)

You are using a "Hybrid Turbojet" whereas I am using a standard "Thermal Turbojet".

The Hybrid Turbojet is supposed to be an upgrade of the Thermal Turbojet that can also operate in rocket-mode, but Fractal_UK may have also given it a higher Thrust/MW than the basic Thermal Turbojet at the same ISP. If that should be the case, then *BOTH* our results are correct- the Thermal Turbojet really does under-perform real life Thermal Rockets like the NERVA (and needs a 47% increase to its Thrust) and the Hybrid Turbojet really is so efficient it's essentially pulling power out of nowhere...

In that case, what needs to be done is the two parts need to be buffed/nerfed until they have more realistic performance. There's no reason the Thermal Turbojet should produce so little Thrust/MW, but there's also no justification for a Hybrid Turbojet pulling Thrust out of thin air...

If the Thrust/MW turns out to be the same for the standard and Hybrid Thermal Turbojets (I'm going to test now to see if this is the case), then I don't know what to tell you- your data simply doesn't match anything I'm getting, and as a mod-developer I'm more inclined to go with my own data if no one else can reproduce your results...

Regards,

Northstar

Share this post


Link to post
Share on other sites

Justification? Well, collecting antimatter in the mod's current state is pretty much pointless, because there's no nice powerful propulsion system to use with it ever since Fractal nerfed thermal rockets. Just for gameplay reasons, a high-thrust high-ISP antimatter propulsion system would be great to have.

Well I have already increased the Isp thermal noozle cap to 5000 and we have already established that 7000 is possible with high tech technology. I intend to make the thermal noozles upgradable with fusion tech from 3000 to 7000 max isp. THis allone should allow you to effectively use Antimatter reactor with significantly higher Isp thermal noozles

Edited by FreeThinker

Share this post


Link to post
Share on other sites

Or rather, the issue seems to be that these reactors are not actually dropping their ThermalPower production when WasteHeat accumulates to levels as low as displayed... The Thermal Turbojet I tested above got roughly the same Thrust as my plane from before, which was operating at 0 WasteHeat and thus had over 849 MW of ThermalPower available:

Actually, the Particle Bed WasteHeat effect on Thermal output is not functional (disabled for upgraded reactors) or only effective at very high waste heat ratio. I intend to change that in the next version where Thermalheat will decrease exponential with wasteheat ratio. This means they will be effective propulsion but when used for electric generation they will quickly suffer from overheating. Which is exactly what I try to accomplish! The only question is how steep it should fall. Right now I use the following formula.

(Note I simplified it here)

CoreTemperature = optimalPebbleTemp + wasteheatRatio^0.25 * (tempZeroPower - optimalPebbleTemp)

MaximumThermalHeat = base.MaximumThermalPower * ((tempZeroPower - CoreTemperature) / (tempZeroPower - optimalPebbleTemp))^ 0.81

Especially the 0.25 constant might need some tweaking as it affects the MaximumTHermalHeat in a big way. For example at WasteheatRatio of 0.1% would result in a CoreTemperature of optimalPebbleTemp + 0.178 times (tempZeroPower - optimalPebbleTemp)

Maximum thermal heat determines how much power you have, for electric generator or thermal nozzles. Thermal engines are then further negatively effected by wasteheat.

Edited by FreeThinker

Share this post


Link to post
Share on other sites

It looks like there is an issue with how Thrust is being calculated with the Thermal Turbojets with reactors that can very their ThermalPower based on WasteHeat accumulation (namely the Particle Bed and Dusty Plasma reactors). With WasteHeat= 0 these reactors still produce the same excessively-low Thrust/MW levels I was observing before...

What basically happens is that when engines is starved from AirFlow, It will demand less thermal heat resulting in less heat production which can now be throthed to 4%. The problem is that the Isp is too low, if it would be higher, it could ask more Heat resulting in more trust and wasteheat.

Internally, it does something funny.


if (myAttachedReactor.shouldScaleDownJetISP ())
{
maxISP = maxISP*2.0f/3.0f;
if (maxISP > 300)
maxISP = maxISP / 2.5f;
}

note shouldScaleDownJetISP is true for all Partcle Bed/ Dusty Plasma reactors.

I need a better way to calculate the Isp of jetengines!

Edited by FreeThinker

Share this post


Link to post
Share on other sites
Well I have already increased the Isp thermal noozle cap to 5000 and we have already established that 7000 is possible with high tech technology. I intend to make the thermal noozles upgradable with fusion tech from 3000 to 7000 max isp. THis allone should allow you to effectively use Antimatter reactor with significantly higher Isp thermal noozles

7000 seconds is still pretty small compared to the 17000 we used to get from those, and the 15500 you can get from the DT Vista. It makes sense that antimatter is more powerful than fusion, right?

Also, 405 gigawatts / (7000 seconds * 9.81 * 0.5) =~ 12 meganewtons, which is the thrust you get from a large fusion reactor burning LFO. It's a little high, and, like dusty-plasma-powered thermal rockets, it begs the question of why you're thermalizing all those nice fast charged particles. As far as endgame tech goes at the tail end of the tech tree, very-high-Isp antimatter rockets make sense, especially given how hard it is to collect large amounts of antimatter.

Share this post


Link to post
Share on other sites

What do you think about an upgraded DTVista as endgame, able to manage antimatter power? Probably is not so realistic, but maybe easier to code and balance... or unbalance...

Share this post


Link to post
Share on other sites
What do you think about an upgraded DTVista as endgame, able to manage antimatter power? Probably is not so realistic, but maybe easier to code and balance... or unbalance...

Nah. The Vista works by a process that wouldn't really work with antimatter - there'd be no point for all those lasers, anyway.

If you're really concerned about game balance, I have another solution: the antimatter reactors in the basic KSPI have a little description, which calls the 1.25-meter one "solid core", the 2.5m one "gas core", and the 3.75m one "plasma core". This has implications for the performance - with the solid or gas core, you'll be driving a thermal rocket and see lower Isp (about what current thermal nozzles can manage), while the plasma core will be the one capable of driving the magnetic nozzle and achieving super-high Isps. I suggest you make two different kinds of antimatter reactors - the gas core ones, which produce all of their output as thermalpower (since the charged particles are instantly thermalized) and don't tweakscale above 3.75 meters or so, and the plasma core ones, which produce most (~80%) of their power as charged particles and don't scale below 3.75 meters. That way, the super-powerful antimatter engines have to be big, meaning they'll be very hard to lift and will use a lot more propellant and fuel to haul around the reactor. This sidesteps the issue in KSP where small rockets keep being overpowered (see the Rockomax 48-7S).

Share this post


Link to post
Share on other sites
7000 seconds is still pretty small compared to the 17000 we used to get from those, and the 15500 you can get from the DT Vista. It makes sense that antimatter is more powerful than fusion, right?

Also, 405 gigawatts / (7000 seconds * 9.81 * 0.5) =~ 12 meganewtons, which is the thrust you get from a large fusion reactor burning LFO. It's a little high, and, like dusty-plasma-powered thermal rockets, it begs the question of why you're thermalizing all those nice fast charged particles. As far as endgame tech goes at the tail end of the tech tree, very-high-Isp antimatter rockets make sense, especially given how hard it is to collect large amounts of antimatter.

Just from a mission-planning perspective, higher Thrust means shorter burn-times. This means that lower-ISP rockets actually will consume less antimatter (but more propellant)- so you have it completely backwards as to which is preferable given the limited availability of antimatter.

Regards,

Northstar

Share this post


Link to post
Share on other sites
Nah. The Vista works by a process that wouldn't really work with antimatter - there'd be no point for all those lasers, anyway.

+1 REP for that succinct, but accurate answer (I'm particularly grateful since my being in the moderation queue means I can't answers questions like this quickly).

If you're really concerned about game balance, I have another solution: the antimatter reactors in the basic KSPI have a little description, which calls the 1.25-meter one "solid core", the 2.5m one "gas core", and the 3.75m one "plasma core". This has implications for the performance - with the solid or gas core, you'll be driving a thermal rocket and see lower Isp (about what current thermal nozzles can manage), while the plasma core will be the one capable of driving the magnetic nozzle and achieving super-high Isps. I suggest you make two different kinds of antimatter reactors - the gas core ones, which produce all of their output as thermalpower (since the charged particles are instantly thermalized) and don't tweakscale above 3.75 meters or so, and the plasma core ones, which produce most (~80%) of their power as charged particles and don't scale below 3.75 meters. That way, the super-powerful antimatter engines have to be big, meaning they'll be very hard to lift and will use a lot more propellant and fuel to haul around the reactor. This sidesteps the issue in KSP where small rockets keep being overpowered (see the Rockomax 48-7S).

At the kind of ISP-increases you're talking about, the plasma core reactors will use less propellant to move payload around, despite the much higher reactor/engine mass. What they will consume more of is antimatter- which is as you noted in very limited supply.

I like the idea of having two different classes of antimatter reactor (I wouldn't go all the way to three- we're already suffering part-bloat), but I'm not sure I like the idea of limiting the size-range of each. Especially given that all rockets in KSP-I are much smaller than in real life, and some players (such as myself) play with Real Solar System- where using TweakScale, Procedural Fairings (especially for the Thrust Plates), and ProceduralParts to get bigger rockets is often desirable...

Also, that reminds me- FreeThinker, the cost-scaling of the reactors using TweakScale is currently very off. In respecting Fractal_UK's original balance, and realism, the larger reactors should be much cheaper/Megawatt than the smaller ones. For instance the 1.25 meter particle bed reactor costs 15K, whereas the 2.5 meter reactor only costs 45K (but produces much more than 3x the power). In TweakScale, like in real life nuclear engineering (where miniaturization is often expensive, and larger reactors are more cost-effective), we need to respect this same balance- the larger reactors should not have cost that scales at the same rate as their size or ThermalPower.

Regards,

Northstar

Share this post


Link to post
Share on other sites

Also, FreeThinker, just to double-check, you're going to include the code for fixing the ISRU refineries to have cryogenic RealFuels modular tanks when RealFuels is installed in one of the next updates, right?

You said you'd include it in 0.8 a while ago- but since you've released several more 0.7.x updates, and I'm starting to wonder whether maybe you forgot about it after NathanKell decided he was going to pull the KSP-Interstellar tank and resource-name fixes from RealFuels as KSP-I Extended has a much more rapid development cycle...

Here the code is again, with a comment I've been including when using it in my current save such as to make the code easier to find/identify (if the fix isn't already in the next update, I suggest including the comment to make it easier for other to identify the purpose of the code in the future...)


//Make ISRU refinery tanks insulated and modular
@PART[FNRefinery]:FOR[RealFuels]
{
MODULE
{
name = ModuleFuelTanks
volume = 1750
type = Cryogenic
}
}
@PART[FNInlineRefinery]:FOR[RealFuels]
{
MODULE
{
name = ModuleFuelTanks
volume = 1750
type = Cryogenic
}
}
@PART[FNInlineRefineryLarge]:FOR[RealFuels]
{
MODULE
{
name = ModuleFuelTanks
volume = 11000
type = Cryogenic
}
}

Regards,

Northstar

Share this post


Link to post
Share on other sites

Also, that reminds me- FreeThinker, the cost-scaling of the reactors using TweakScale is currently very off. In respecting Fractal_UK's original balance, and realism, the larger reactors should be much cheaper/Megawatt than the smaller ones. For instance the 1.25 meter particle bed reactor costs 15K, whereas the 2.5 meter reactor only costs 45K (but produces much more than 3x the power). In TweakScale, like in real life nuclear engineering (where miniaturization is often expensive, and larger reactors are more cost-effective), we need to respect this same balance- the larger reactors should not have cost that scales at the same rate as their size or ThermalPower.

Regarding, cost I'm leaning towards a ScaleFactor 2 (surface area) which I think is generous enough considering the overal increased power.

- - - Updated - - -

As for numbers, I can probably help a little more with that.

Project Rho's page on Exotic Weapons has a bit about the kinds of things antimatter gets up to during annihilation. Based on what I found there, it looks like approximately 55% of the energy from an antimatter reaction will show up as charged pions, while the rest is going to end up as gamma rays (which, for the sake of simplicity, I think the mod refers to as thermal power). Of course, there are other small details to consider as well - mainly, you're not going to thermalize all the gamma rays. What that means is that, say, the 405-gigawatt antimatter reactor, in order to produce 405 gigawatts of extractable power, might have to annihilate considerably more than 405 gigawatts' worth of antimatter. The gammas, particularly those formed as a result of neutral pion decay, probably are going to be tricky to trap. Project Rho seems to be of the impression that about 1/3 of the energy of the reaction will be from neutral pions, so if we assume that half the neutral pions and their subsequent decay-product gamma rays escape the reactor without providing useful heat, that means that only ~66% of the antimatter energy is captured. Furthermore, if we assume that all of the charged pions are captured, that means that our charged particle ratio relative to the total power is going to be something like 0.55/0.66, or about 83%. Assuming that from that point on, the charged pions are expelled from the magnetic nozzle with near-perfect efficiency (an assumption that the mod apparently does make for the other kinds of charged particles), then we'll see quite handsome engine performance (in the neighborhood of 320 gigawatts, if you make the same assumptions I did.)

As of now, antimatter reactor consumption is a little strange. If we assume that the speed of light c is the same in KSP as it is for our universe, the 15 milligrams/second draw of the 405 GW reactor comes out to E=2*0.000015*299792458^2 or about 2.7 terawatts. That's a rather low efficiency, about 15%.

However, that's forgetting one small detail: For the Alcubierre drives contained in the mod, the speed of light for the itty-bitty kerbal universe is assumed to be 1/10 of the one in our universe. Originally, this was just to keep the drives from being unreasonably fast relative to the system, but it also has the implication of reduced energy/mass equivalence. Using c*0.1 for our antimatter yield gives us E=2*0.000015*29979245.8^2, or only about 27 gigawatts! In short, that antimatter reactor is getting a lot of power out of nowhere if you assume that the speed of light is consistent between the antimatter reactors and the Alcubierre drives.

To find the antimatter draw that would really be required for a 405 gigawatt reactor assuming the above 66% total efficiency, we could use M = E/2*c^2, so M = (405 GW/0.66 eff) / (2*29979245.8^2) = 0.000341 kg/s or 341 milligrams per second. It's a little high, but I think it's reasonable - if it's really an issue of game balance, the capacity of the antimatter containers could always be bumped up a bit.

Sorry, I a bit confused about the numbers. You say that 83% of all usefull energy from antimatter reactor will be in the form of charged particles which can be used for magnetic noozles, correct?

Edited by FreeThinker

Share this post


Link to post
Share on other sites
Just from a mission-planning perspective, higher Thrust means shorter burn-times. This means that lower-ISP rockets actually will consume less antimatter (but more propellant)- so you have it completely backwards as to which is preferable given the limited availability of antimatter.

Regards,

Northstar

Yeah, I'm aware how rare antimatter is in real life, but there's a huge amount of it in KSP - you just have to fly a big collector array out to Jool, and you can get kilos of the stuff in a handful of years.

When KSPI got its magnetic nozzle nerf in the first place, I tried adding in my own config for an antimatter "torch drive," which had a thrust of 3750 kN, an Isp of 56000 seconds, and antimatter consumption directly proportional to that thrust power (I assumed real-world c for my mass/energy calculations). I don't know what you'd get out of a 405 gigawatt antimatter reactor (the thrust power of my drive was over a terawatt because I wanted >1G acceleration) but the ratio of antimatter to propellant worked very well - some very large ships needed two of the biggest bottles, but collecting that much isn't that prohibitively difficult with endgame tech. Sure, there comes a point when your Isp is too high, and you'd get a more effective interplanetary ship by gearing it down, but 56000 seconds isn't nearly there yet. For a drive using a 405-gigawatt reactor (assuming 80% charged particles), I'd probably cut the Isp slightly to about 42000 seconds (after all, we're probably not trying to make actual torchships here the way I was), which would then give us a very reasonable maximum thrust of roughly 1500 kN. It would be like running a much stronger DT Vista, which I think should be the goal. If I have some time later today, I'll see if I can mess with the reactor configs and get this to work myself.

Sorry, I a bit confused about the numbers. Summorised you say that 83% of all usefull energy from antimatter reactor will be in the form of charged particles which can be used for magnetic noozles, correct?

Yeah, the ~80% range looks about right. It could easily be slightly higher or slightly lower than that, depending on how efficient the reactor is at capturing gamma rays, but just some quick-and-dirty back-of-the-envelope calculations suggest that a reasonable efficiency for that might give you an 80% charged particle ratio.

Share this post


Link to post
Share on other sites

Yesterday, I did a some experimenting with a combining the Thermal Trusters with hermal Jet Engines. Allthough Thermal Jet Engines attached to Particle Reactor are very powerfull in the lower athmosphere, the low performace in high athmosphere, makes getting into sub orbit without a steep climb very difficult. Getting anywhere further than low ortbit and the dead weight of the unused nuclear reactor becomes counter productive. This effectivly limits the (unupgraded) Thermal Jet Engine to atmospheric/suborbital exploration only.

I think the main problem is that the reactors can only be connected to either a thermal thruster or thermal jet engine, not both at the same time. That's because the heat from the reactor needs to be efficiently transported to the thermal thruster or thermal jet engine. For most reactors this makes a lot of sense, except for Molten Salt Reactors, which thermal heat (in the form of molten salt) could in theory be transfered further. My Idea is to allow Molten Salt reactors to transfer their reactor heat one part further than other reactors. This would allow you to connect a single Molten Salt reactor to both thermal truster and thermal jet engines at the same time. Another advantage it it would allow you to put the heavy Salt core reactor in a more central location of the vessal, which should be very usefull for any airplaine. This will give the Molten Salt reactors, which are significantly less powerfull compaired to the Particle Reactor, an edge when building SSTO as they allow less dead weight to be carried around and located at a more convenant location.

This valuable feature should be easily implemented by slightly changing the method in how the thermal thruster/jet searches/thermal generator finds a thermal source, I allow it to look 1 part futher. I Now only need to deal with dividing the thermal heat over multiple consumers evenly.

Edit: now that I think about it, perhaps the same could be done with charged particles. This would allow reactors which generate charged particles to used both thermal trusters/jet and magnetic noozles at the same time.

Edited by FreeThinker

Share this post


Link to post
Share on other sites
Yeah, the ~80% range looks about right. It could easily be slightly higher or slightly lower than that, depending on how efficient the reactor is at capturing gamma rays, but just some quick-and-dirty back-of-the-envelope calculations suggest that a reasonable efficiency for that might give you an 80% charged particle ratio.

I wonder if we could theoreticly convert charged particles back to thermal heat. Preferably this would be changable by the click of a button or slider. Combined with my idea described above this would allow you to switch between Thermal truster low ISP/high trust mode, to Magnetic Noozle High ISP/Low trust mode. In essence the best of both worlds

- - - Updated - - -

Here the code is again, with a comment I've been including when using it in my current save such as to make the code easier to find/identify (if the fix isn't already in the next update, I suggest including the comment to make it easier for other to identify the purpose of the code in the future...)


//Make ISRU refinery tanks insulated and modular
@PART[FNRefinery]:FOR[RealFuels]
{
MODULE
{
name = ModuleFuelTanks
volume = 1750
type = Cryogenic
}
}
@PART[FNInlineRefinery]:FOR[RealFuels]
{
MODULE
{
name = ModuleFuelTanks
volume = 1750
type = Cryogenic
}
}
@PART[FNInlineRefineryLarge]:FOR[RealFuels]
{
MODULE
{
name = ModuleFuelTanks
volume = 11000
type = Cryogenic
}
}

Regards,

Northstar

Ah, I must have forgotten .. I will add it with a small alteration by replacing :FOR[RealFuels] by :NEEDS[RealFuels]:FOR[WarpPlugin]

Edited by FreeThinker

Share this post


Link to post
Share on other sites
I wonder if we could theoreticly convert charged particles back to thermal heat. Preferably this would be changable by the click of a button or slider. Combined with my idea described above this would allow you to switch between Thermal truster low ISP/high trust mode, to Magnetic Noozle High ISP/Low trust mode. In essence the best of both worlds

I thought it was already established that unused charged particles are automatically thermalized by the reactor? This was consistent with my observations with the thermal jets - the jets were putting out about as much power as the thermalpower + chargedparticle components of the power together, not just thermalpower. I believe antimatter reactors with charged particles would behave about the same with thermal rockets as they do now (which is still very buggy, by the way).

Share this post


Link to post
Share on other sites
I believe antimatter reactors with charged particles would behave about the same with thermal rockets as they do now (which is still very buggy, by the way).

Buggy, How so?

Share this post


Link to post
Share on other sites

I tested the large 405 GW reactor with a thermal nozzle and liquid fuel, and it gave me something like 750 kN at 3000 seconds (with the reactor running at 100%). So essentially, almost all the reactor power is vanishing into thin air, without showing up as useful work or even waste heat.

Also note that even if the energy was conserved and 80% efficiency was assumed, 3000 seconds at that thrust power would give around 22 meganewtons (22,000 kN) of thrust. An Isp of 5000 seconds would give 13,200 kN, and even if you drove that up to 7000 seconds, you're still looking at over 9,000 kN. While not exactly game-breakingly strong, that's still enough to accelerate at a quite ludicrous number of Gs for even most large spacecraft. Ergo, mega-powerful antimatter engines really need a higher Isp to make them sane - There's just no need for an engine that thrusty, and even if there is, it probably involves partcounts bigger than most computers can handle.

Edit: Derp. I had bad parens in all that math, and it was totally bogus. Should be fixed now. :blush:

Edited by GreeningGalaxy

Share this post


Link to post
Share on other sites

9bDRRE6.jpg

Jeb in his Nuclear Molten Salt Powerered Thermal Jet. Notice how the heavy Molten Salt reactor is positioned in the center, keeping the aircraft in perfect balance, forever.

Edited by FreeThinker

Share this post


Link to post
Share on other sites
http://i.imgur.com/9bDRRE6.jpg

Jeb in his Nuclear Molten Salt Powerered Thermal Jet. Notice how the Molten Salt reactor is positioned in the center, keeping the aircraft in perfect balance.

Edit: I have some trouble detecting parts in surface attached nodes, does anyone know a mod mod which looks for specific modules in surface attached parts?

1) wait, you have them not directly attached?

2) I do not know - I would like that functionality as well, as I have mentioned in a previous post.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.