Rethinking the nuke engine, where can we take it from here?

[Jovus]

Eh... I prefer to think that we're getting massive overheating because we're running kerosene through the engine and getting LHyd isps out of it. Unrealistic, but this is #LOLSOKERBAL-land to begin with and there might as well be a trade-off and interesting mechanic for the heating system aside from reentry.

Unrealistic? Nah. It's a gas core engine that they're running RP-1 through. It gets ISP similar to solid core NERVAs because of the massively increased core temperature (ISP also being proportional to the square root of the exhaust temp, naturally). Why are they doing this? Because Kerbals haven't discovered hydrogen yet. (Water is insufficiently explosive for them to be particularly interested.)

So yes, the LV-N overheats, because in order to get the required ISP it runs at higher than designed temperature. They probably found the specs somewhere on the Mun; possibly near the Neil Armstrong memorial.

Yes, this is a joke post.

[Laie]

That is the issue, with no radiator parts there is no solution.

What regex said. Also, do we even need to have this problem?

First we create a part that produces insane amounts of heat, then we create a part that drains insane amounts of heat. What's the point? Can't we just integrate the heatsink into the engine, increase it's mass by another 500kg, and call it a day?

More generally... I understand that heat managment is important IRL. I also understand that KSP is a game and doesn't need to depict everything 100% accurately. But still, is it necessary to go so completely overboard, invent heat sources out of thin air just so the player has to care about heat managment?

[Caelib]

Even when you exclude the heating issue from the problem, I don't think the nuclear engine is really that useful anymore. Maybe the fact that there aren't any large 2.5m liquid-only tanks is part of the problem because you end up carrying so much dead weight from half-empty tanks from Oxidizer drained compartments.

Edited May 8, 2015 by Caelib

[sephirotic]

I really wished that Squad would do A PROPER implementation of tweaks for stock tanks. Making the LV-N burn only liquidfuel without allowing to customize a tank to use ALL OF ITS VOLUME with a specific fuel type is just, stupid IMO.

The rebalance of the engine by increasing it's weight to 3 tons is also pointless, the LV-N were already nerfed in the first place in regard to Real Life NTR's wich had a much higher TWR.

Proposions to use Nuclear engines in real life always treats then as ferrys and never as lifter engines, engine fraction is very important and having a 3 ton engines for interplanetary travel is just stupid.

For the LV-N to be usefull again we really need to have it's weight brought down again for fuel fraction sake, and the fuel tanks to be twekable.

I would really also REALLY would want to se the LV-Ns as a trimodal engine with a proper permanent eletrical generator, and a switchable mode for Lox Augmented Thrust burning Oxidizer with some proper thrust.

And a 0.625 nuclear engine, a MITEE analogue with about 10kn of thrust for 0.25 tons of weight would also be very interesting for a lower engine fraction interplanetary vessel.s

[Gaarst]

Even when you exclude the heating issue from the problem, I don't think the nuclear engine is really that useful anymore. Maybe the fact that there aren't any large 2.5m liquid-only tanks is part of the problem because you end up carrying so much dead weight from half-empty tanks from Oxidizer drained compartments.

I could get 8 km/s of delta/v from a 25t nuclear powered ship with a 5 t payload using 6 Mk2 jet fuel tank and the stack tri-coupler, the 6 fuel tanks approximately have the size of a Jumbo-64, height and width.

So the LV-N engine can still be used with a little imagination.

[Renegrade]

Eh... I prefer to think that we're getting massive overheating because we're running kerosene through the engine and getting LHyd isps out of it. Unrealistic, but this is #LOLSOKERBAL-land to begin with and there might as well be a trade-off and interesting mechanic for the heating system aside from reentry.

Yeah - the established nerfs to it are more-or-less adequate. My point is largely that a real life NTR isn't some magic bullet panacea BS, and that the curveballs Squad is throwing at it are plenty well justified, even if they're weird, and thus I oppose their removal without the addition of other counter-balancing points.

Lulz at the thought of the forum response to actual hydrogen tankage. They should all play RealFuels at least once (i've done so twice already and will play with it again soon).

The problem I see is that we already have enough part clutter. I think some degree of proceduralising will become inevitable: same basic shape in different sizes for the x-to-y adapters, four lengths for the cylinders. Color-coded by content. And if you want an orange tank full of xenon, so be it. That will be 2 million funds, please.

I'm not opposed to being able to repurpose tanks on a right click, but they gotta obey the mass ratios established in 1.0 (which seem to be relatively consistent, finally). Clicking a tank over to pure LF shouldn't fill it at a 9:1 ratio.

For gameplay purposes, I'm strictly against boiloff. Like life support or RemoteTech, it shouldn't be tossed at everyone.

While I agree about the RemoteTech point (once you've mastered RT2 operations, they just become drudgery. It's only fun when you're setting up the network, doing stationkeeping is a PITA), I disagree about the LS point. It makes for some nice gating (although I wouldn't want something as detailed as TAC_LS).

If Squad feels that this extraordinary engine needs an extraordinary nerf, I'd like to repeat Brotoro's idea of making nukes spool, much like jets do. Let them take a few moments to come to full thrust, and again some time to throttle down. That would even be kinda-sorta realistic, while overheating definitely is not.

Enabling/disabling an engine bypasses the spooling entirely, FYI. I tried that out back in 0.90 (and you can't specify them separately back then either. Haven't tried since). Also don't be too absolutist on NTR designs and their drawbacks - nobody has EVER flown one.

The rebalance of the engine by increasing it's weight to 3 tons is also pointless, the LV-N were already nerfed in the first place in regard to Real Life NTR's wich had a much higher TWR.

Actually, I wrote a long and detailed analysis of that - and based on comparisons with other KSP engines, the old LV-N is actually like a well-made second generation NTR, which has never even physically existed. (Only first generation NERVA-types were even partly built, the follow-up designs during SDI never progressed beyond paper studies).

NERVA itself massed almost as much as an F-1 and had a third of the thrust of a J-2.

Keep in mind that rocket TWRs in KSP are about a sixth to a third of what they are in real life for balance reasons.

Proposions to use Nuclear engines in real life always treats then as ferrys and never as lifter engines, engine fraction is very important and having a 3 ton engines for interplanetary travel is just stupid.

You mean like the Saturn C5-N?

A hydrogen powered engine would be rather hard to bring on any sort of interplanetary trip when it's fuel would evaporate long before you left the Earth's SOI.

I would really also REALLY would want to se the LV-Ns as a trimodal engine with a proper permanent eletrical generator, and a switchable mode for Lox Augmented Thrust burning Oxidizer with some proper thrust.

No, bad! It's bad enough that it has second generation stats, it doesn't need third generation capabilities. Not unless the Kerbol system was a lot bigger, and the tech tree a lot longer. Mind you, I wouldn't mind seeing an ACTUAL Pluto-equivalent (Eeloo is inside Saturn's orbit in relative terms), AND also having a longer tech tree.

[NathanKell]

The NTR has always had a higher (relative) TWR than other engines, as Renegrade points out. Indeed, the old NTR had about spot-on TWR for a 60s NTR, whereas the other engines have not just 1/3 to 1/6th, but more like 1/4 to 1/8th the TWR of even 1960s counterparts.

As for what LF and Oxy are, I suggest you look at their specific heats in Resources.cfg. The new Isps make sense for those specific heats (max of 295-300 staged).

[sephirotic]

Actually, I wrote a long and detailed analysis of that - and based on comparisons with other KSP engines, the old LV-N is actually like a well-made second generation NTR, which has never even physically existed. (Only first generation NERVA-types were even partly built, the follow-up designs during SDI never progressed beyond paper studies).

NERVA itself massed almost as much as an F-1 and had a third of the thrust of a J-2.

Actually the original NERVA that was tested back in the 60's weighted much more than the F1 at a massively 34 tons. But It also had a very high thrust.

You mean like the Saturn C5-N?

No, I see Kerbal's NTR as more in line with the Soviet variant Stat-wise, (even tough it is visually more similar to the original NERVA), the RD0410, a smaller, much lighter, with a better TWR and higher efficiency engine more suitable for High stage and interplanetary burns, It only weighted 2.0 tons while having 36KN of thrust and 910 of VISP. So the kerbal LV-N is not much more nerfed compared to the real life counterpart, you sacrifice a little of ISP (910 vs 800) for a little more thrust: (18 vs 26 TWR with the old LV-N stats).

Keep in mind that rocket TWRs in KSP are about a sixth to a third of what they are in real life for balance reasons.

But JET, ION engines and even RCS pods are buffed and have BETTER TWR ratio than real life counterparts. The point of nerfing the TWR (and ISP) of engines in KSP is to compensate for the much smaller orbital velocity needed to escape Kerbin's surface, but once you are free in space, the orbital and interplanetary engines should be actually buffed in relation to real life engines as not to make the burns ridiculously long, that is why the ION engine doesn´t have an abismal 50 newtons of thrust (not KILOnewtons).

I don´t see a point in making the LV-N excessive heavy and sacrificing engine fraction for interplanetary stages.

No, bad! It's bad enough that it has second generation stats, it doesn't need third generation capabilities. Not unless the Kerbol system was a lot bigger, and the tech tree a lot longer. Mind you, I wouldn't mind seeing an ACTUAL Pluto-equivalent (Eeloo is inside Saturn's orbit in relative terms), AND also having a longer tech tree.

I don´t see why it's bad, we are not talking about higher specific impulse here wich would make then trully cheaty, just about energy production (I don´t believe we don´t have a proper reactor apart from fuel cells now that we have such power hungry parts like the scientific LAB and the INSITU converter), and the extra thrust don´t make the designs of rockets smaller or easier to make, just make the burns smaller. It's about reducing GRINDING not about making the game easier.

If the point was to nerf the original LV-N TWR, than I'd rather had it's thrust lowered instead of the weight increased, again, I really wanted a MITEE equivalent in KSP, it's all about engine fraction!

Edited May 8, 2015 by sephirotic

[Aerindel]

So the take away is that in reality there really is no practical way to accomplish interplanetary travel.

[Brotoro]

So the take away is that in reality there really is no practical way to accomplish interplanetary travel.

Umm. No. Do you mean in real life, or in KSP? Either way, the answer is no. There are no technical reasons we couldn't do interplanetary travel in real life, and it certainly is doable in KSP.

[regex]

So the take away is that in reality there really is no practical way to accomplish interplanetary travel.

What are you talking about? The LV-N works perfectly fine, people just need to relearn how to use it, and use it appropriately. If it's so terrible use a Poodle or 909 for a transfer engine, it's not like it's hard to bring enough fuel to orbit to get 6km/s out of them... A transfer to Duna costs less than 2km/s ferchrisakes...

[Brotoro]

As for what LF and Oxy are, I suggest you look at their specific heats in Resources.cfg. The new Isps make sense for those specific heats (max of 295-300 staged).

I can't find such a file. Where is it hidden?

[Aerindel]

Real life. If atomic engines require giant tanks of hydrogen that would boil away before you could use it and other engines have such low ISP that you would need monstrous tanks to move very small payloads it seems that there is no practical way to send much more than a couple of guys at time to another planet without a major break through in physics and the only reason LV-N's work is because their fuel source is unrealistic.

[Brotoro]

You can keep liquid hydrogen in a tank for long periods if you bring along the necessary refrigeration equipment.

You can also run nuclear engines on other fuels besides hydrogen (albeit at lower Isp).

[regex]

I can't find such a file. Where is it hidden?

It's GameData/Squad/Resources/Resources.cfg

[Brotoro]

In GameData/Squad/Resources/ I see ResourceDefaults.cfg and ResourcesGeneric.cfg

Ah! It's apparently ResourcesGeneric.cfg...

RESOURCE_DEFINITION

{

name = LiquidFuel

density = 0.005

unitCost = 0.8

hsp = 2010

flowMode = STACK_PRIORITY_SEARCH

transfer = PUMP

isTweakable = true

}

RESOURCE_DEFINITION

{

name = Oxidizer

density = 0.005

unitCost = 0.18

hsp = 1551

flowMode = STACK_PRIORITY_SEARCH

transfer = PUMP

isTweakable = true

}

[regex]

In GameData/Squad/Resources/ I see ResourceDefaults.cfg and ResourcesGeneric.cfg

Which one has a bunch of resources in it?

[Laie]

I could get 8 km/s of delta/v from a 25t nuclear powered ship with a 5 t payload using 6 Mk2 jet fuel tank and the stack tri-coupler,

Other scenario: One half-size Mk3 tank, six nervas. Can push an orange tank to Duna and back, at a respectable 4m/s² (5-Minute transfer burn). If it wasn't for the heat...

Enabling/disabling an engine bypasses the spooling entirely, FYI.

Didn't think of that. But AFAIK this only buys you instant shutdown. The engine spools down in the background, if you reactivate it after a few seconds you have appropriately reduced thrust and spool to up again.

So the take away is that in reality there really is no practical way to accomplish interplanetary travel.

Relatively speaking, it's no more impractical than using Poodles in KSP. Which ain't bad, but... Remember the tug at the beginning of this post, using 18t of LV-N for 25t of fuel? Lots of people here would cringe at the thought of adding so much dead mass. But that contraption still has slightly better TWR, and considerably better delta-V, than the same amount of fuel with a single Poodle.

Yeah - the established nerfs to it are more-or-less adequate. My point is largely that a real life NTR isn't some magic bullet panacea BS, and that the curveballs Squad is throwing at it are plenty well justified, even if they're weird, and thus I oppose their removal without the addition of other counter-balancing points.

I reluctantly agree. But what bugs me is that everything Squad does seems to increase part count more than anything else. It's already a small and weak engine, requiring multicouplers or structural fuselages if you want any kind of playable TWR. This also requires some ingenuity if you want to mount another stage below. Now heat, which is manageable, but drives up the part count even more. Whether the vessel's TWR is goddamn low, or the clock is in the yellow: either way, Nervas require patience. That is a mighty nerf and I hate it.

[sephirotic]

Real life. If atomic engines require giant tanks of hydrogen that would boil away before you could use it and other engines have such low ISP that you would need monstrous tanks to move very small payloads it seems that there is no practical way to send much more than a couple of guys at time to another planet without a major break through in physics and the only reason LV-N's work is because their fuel source is unrealistic.

That is why different fuel types have been throughly proposed in the last decades of theoretical researches for Nuclear thermal rocket.

Particularly Anmonia, Methane and Even Salt Water. More interesting, is the LANTR mode too.

[NathanKell]

Sorry, I meant ResourcesGeneric. Oops. :]

[Brotoro]

Sorry, I meant ResourcesGeneric. Oops. :]

No problem. It just wasn't coming up when I searched for that filename.

Also, I can't find a match for an oxidizer with that specific heat value...

[Archgeek]

Other scenario: One half-size Mk3 tank, six nervas. Can push an orange tank to Duna and back, at a respectable 4m/s² (5-Minute transfer burn). If it wasn't for the heat...

Have you tried one half-size Kerbodyne tank, partially drained of oxidizer and with a Rhino engine on the back of it, flanked by six stacks of two mk1 LF tanks with a NERVA? The rhino and oxidizer can be used to boost TWR when needed, and when off makes a disturbingly effective radiator of the heat flowing into the central tank from the radial ones. Testing showed flux numbers of over 50k (and this was in 1.0 when things were running dang hot). In 1.02 the assembly runs out of gas first, often sun-diving or going out past Eeloo or hyperbolically away from Kerbol.

[Renegrade]

Didn't think of that. But AFAIK this only buys you instant shutdown. The engine spools down in the background, if you reactivate it after a few seconds you have appropriately reduced thrust and spool to up again.

That is indeed correct. Gettin' it running again can be a bit pokey, so you gotta be careful. I was able to successfully land with a modified LV-N like that in my tests.

I reluctantly agree. But what bugs me is that everything Squad does seems to increase part count more than anything else.

True, and I do consider keeping part counts manageable to be an important goal. The array-of-wings thing for instance did not appeal to me. The new heat seems to be fairly manageable though, my old "nerv" test craft pretty much had to do periapsis kicks to get to the Mun originally, but had no problems when I launched it again in 1.0.2 (it had only four of those cheapy fins that don't seem to be very good radiators. They got rather warm but the craft was able to burn 860 m/s without meltin' 'em).

I'd actually like to see radiators and heat management (ala Interstellar but uh, a bit less clunky and poorly documented~) become a thing in KSP.

[Gravaar]

Real life. If atomic engines require giant tanks of hydrogen that would boil away before you could use it and other engines have such low ISP that you would need monstrous tanks to move very small payloads it seems that there is no practical way to send much more than a couple of guys at time to another planet without a major break through in physics and the only reason LV-N's work is because their fuel source is unrealistic.

http://en.wikipedia.org/wiki/Project_Orion_%28nuclear_propulsion%29

Would work just a little upsetting to the people left behind

[Temeter]

http://en.wikipedia.org/wiki/Project_Orion_%28nuclear_propulsion%29

Would work just a little upsetting to the people left behind

Ah, the Deus Ex Nukina kind of scientifical study.