Realistic NERVA engines?

[Gilliam]

Heyho guys, long time forum lurker but first time poster here. After having discovered KSP from scott manley's channel this summer, i've absolutely loved to play it. I am thinking of contributing to this awesome community with my ship designs, and perhaps opening a youtube channel dedicated to KSP playthroughs. But anyway, enough about me, lets go on to the topic!

So I noticed a couple of things after reading the wikipedia page on the NERVA (Nuclear Engine for Rocket Vehicle Application) late 60's / early 70's NASA research project:

1. Holy guacamole, why arent these things in use today? they can realisticly reach 800-1100 seconds of specific impulse. I guess its because of the "N word", but the risk seems limited seeing as these engines are wrapped in carbonfibre and heat resistant composites. Land based nuclear applications seem a lot more risky to me, considering earthquakes and tsunami's and all that (COUGHfukushimaCOUGH).
   
   
2. Kerbal space program is mentioned in the "in fiction and pop culture" section. Yay KSP!
   
   
3. In comparisson to reality, the 60kN thrust of the LV-N atomic rocket motor in KSP is severely underpowered. I understand this from a game balance standpoint, but cant the game be balanced in a more realistic way? I mean, the last prototype NERVA engine produced 334 kN of thrust at 380 S isp (atmosphere) and with 850 S isp in vacuum.
   
   
4. Just like in game, NERVA engines in real life also have issues with overheating. Contrary to the KSP mechanics however, the thrust and fuel efficiency of real NERVA engines increases with temperature with no theoretical limit. This makes it a careful balancing act of reaching high temperatures in the fuel expansion chamber, without causing the whole thing to melt or explode.
   
   

Now, how can the LV-N atomic rocket motor better resemble the real thing without being overpowered? Well, i thought of the following possibility:

• Match the thrust and isp of the engine with the NASA prototype (so 334 kN and 850 S isp).
  
  
• Make the engine thrust drop to 150 kN at 380 S isp at sea level. This matches reality since rocket engines dont only lose isp, but also thrust inside atmospheres.
  
  
• Give the LV-N an even stronger tendency for overheating, especially inside the atmosphere. It could for example only function at 50% thrust at kerbin sea level without exploding.
  
  
• Make the thrust of the LV-N rely on its level of overheating. So, at 0% overheating it produces no thrust. At 80% overheating it would produce 80% of its maximum thrust.
  
  

In summary, this would make the stable thrust of the LV-N at sea level:

150 kN * 50% * 80% = 60 kN

It may seem a little complicated, but thats about how I understand these engines to work with real world physics. The whole change i described above would make LV-N's just as crappy at sea level as they are in KSP now. But it will give them progressively better thrust at higher altitudes just like the real thing.

P.S. NERVA engines also dont use any liquid oxygen for thrust, since there are no chemical reactions involved. However, I dont know whether this is possible to balance in KSP.

P.P.S. This could also give the opportunity to add a new module to KSP, the "Cooling cycle" or "Heat dissipator". This could dissipate a certain amount of heat from engines on your ship, allowing you to run them at higher thrusts without exploding. In order for these to be balanced, they should be progressively weaker as you slap more engines onto your ship. Furthermore, the module could be quite heavy or a strong electricity drain. Or it could require its own reserve of coolant (like liquid helium).

P.P.P.S. I also read something obscure about a NASA project called LA-NTR. Im assuming this stands for "Liquid oxygen Assisted Nuclear Thermal Rocket". They function the same as NERVA engines, but they have the possibility of injecting liquid Oxygen as a sort of "afterburner" effect. This tripples thrust, but drasticly reduces fuel efficiency. This type of engine could be interesting in KSP aswell.

Im really curious as to what you guys think of these possibilities. Do they fit within KSP, and do they give us mad kerbal rocketeers some more awesomeness to play with? Or are they just overcomplicating things?

Cheers and happy space trucking!

[regex]

Given that the solar system is 1/10th the size of our actual solar system, I'd say KSP's nuclear engine works great and needs no improvement. If you don't think so, just edit the part config file to match the values you think it should have.

[Umlüx]

such a nerva would be severely overpowered in the kerbal universe. i think they are just fine the way they are.

if you need more thrust, add more of them

[Sirrobert]

Gameplay ballance and Realism are 2 terms that never work well tougether. You shouldn't try mixing them

The Atomic motor in game is great. High ISP at the cost of low thrust. 300KN thrust is more than most atmospheric engines we have. Why would you use anything else if you gave a NERVA that much?

[Gilliam]

Thanks for responding and i respect your opinions regex. But I wonder, don't you find it frustrating to have to do 20+ minute burns to get larger sized ships from kerbin to a Jool transfer orbit? I sure do! Now, if this would match reality I wouldnt be bothered, but real nuclear engines are a lot better in space then the KSP counterparts (as described). So why should that element of frustration exist? Also, I'm not editing KSP files because i consider it a game and not an exercise in computer programming .

Regarding game balance, an engine can be balanced in different ways then just the thrust and ISP. My example of making it overheat faster, and making it require heavy cooling equipment is one way. Another way would be to just increase the mass of the engine itself.

Also, what do you think of the other propositions that i made, like varying thrust based on atmospheric density and engine overheating?

And what do you think about adding a cooling unit and / or a LA-NTR based rocket engine to the game?

[Tw1]

You make some good points about the heating.There would need to be a reliable way to build it up though.

All KSP engines vary in isp without varying in thrust, pretty unrealistic, yeah. There was a mod out there that changed that, but it seemed out of date, so I never tried it.

[MalfunctionM1Ke]

[*]Holy guacamole, why arent these things in use today? they can realisticly reach 800-1100 seconds of specific impulse.

Because Humans turn smallish, green and keep screaming all the time if they use them too often

BTW there is a nice article on Wikipedia which I discovered recently. Containing all types of engines with pro's and con's.

http://en.wikipedia.org/wiki/Rocket_engine

Edited November 5, 2013 by MalfunctionM1Ke

[Psycix]

The NERVA can not be realistic in KSP for gameplay reasons. It would be massively overpowered.

You may say: Then isn't it overpowered in real life as well?

Yes it is.

Nuclear rockets are much much better (in terms of performance) than conventional rockets.

Edited November 10, 2013 by Specialist290
Edited to remove off-topic content.

[MalfunctionM1Ke]

There are probably just 1 or 2 Nerva Engines landing in your backyard in your entire lifetime..

Nothing to be too worried about

[Gilliam]

Well, I agree with the concern that the LV-N engine would simply be too overpowered to use anything else. However, I wanted to suggest that there are other ways of balancing an engine, other then just giving it a really low thrust.

One way to do it is to make the engine overheat way quicker, forcing you to adjust thrust to lower levels. A new cooling unit could help with this, but making this very heavy would balance that out aswell. Another option is to increase the mass of the engine itself. As far as I can tell, real nuclear thermal engines would way a lot more then 2.25 tonnes, since their basicly solid chunks of Carbon, Uranium and heat shielding materials.

Also, the need for your engine to "spool up" as its level of overheating increases also helps balancing. This is because you'd essentially waste fuel while your cold engine isnt providing any thrust.

I just think that these changes would make nuclear based vehicles a lot more interesting, and have a lot more depth to them.

Also, what do you think about a separate LA-NTR engine: a nuclear engine that has similar thrust and ISP to the in game version, but can inject liquid oxygen for a higher thrust / lower ISP afterburner effect?

[Anton P. Nym]

Nuclear rockets are much much better (in terms of performance) than conventional rockets.

Um, no. I used to agree with your point, but some digging a while ago showed me that NERVA wasn't nearly as "solved" a problem as I thought. The killer, in a literal sense, flaw is the radiation containment; full shielding would've been too massive, so designs just used a "shadow" shield that only worked to protect the craft above the engine. Anything outside of a roughly 60° cone of dead-on front would be exposed to the reactor.

So every time in KSP that you turn your LV-N craft retrograde to another craft within, say, ~500m in real life you'd be scrambling that craft's electronics and giving its crew (if any) cancer.

I'd love to see a "tamer" version of NERVA someday, but I certainly understand the opposition to its use.

-- Steve

Edited November 10, 2013 by Specialist290
Edited to remove content from quote that could have been construed as inflammatory.

[Sirrobert]

Um, no. I used to agree with your point, but some digging a while ago showed me that NERVA wasn't nearly as "solved" a problem as I thought. The killer, in a literal sense, flaw is the radiation containment; full shielding would've been too massive, so designs just used a "shadow" shield that only worked to protect the craft above the engine. Anything outside of a roughly 60° cone of dead-on front would be exposed to the reactor.

So every time in KSP that you turn your LV-N craft retrograde to another craft within, say, ~500m in real life you'd be scrambling that craft's electronics and giving its crew (if any) cancer.

I'd love to see a "tamer" version of NERVA someday, but I certainly understand the opposition to its use.

-- Steve

When is a rocket EVER going to fire it's engine while pointing said engine at another craft while 500m away?

And radiation isn't a factor in outer space anyway. There is already radiation out there

[jwenting]

that's an open cycle engine. A closed cycle engine would be heavier, but wouldn't spew radiation all over the place (the shielding problem remains of course, adding enough shielding to shield the entire 360 around the engine would likely be prohibitively heavy).

An open cycle engine was actually test fired several times in the USA as part of a program to develop a nuclear powered nuclear tipped supersonic, intercontinental, cruise missile.

The project was eventually abandoned in favour of ICBMs when no good way to test fly the craft could be thought up, given that the flight profile would end with the missile crashing into something, breaking apart, and spreading radioactive waste over a large area. One idea was to tether it to a long steel cable connected on the other end to a very long pole rammed into the sea floor somewhere far out in the Pacific, but that was dismissed as being impractical (not for environmental concerns primarily, but because it would be harder to get the pole in place than to actually build the missile).

[numerobis]

And radiation isn't a factor in outer space anyway. There is already radiation out there

There is radiation, but adding lots more exposure means astronauts will roast even faster than expected on their interplanetary trip. Also, the reactor is constant exposure, whereas deep-space radiation comes largely in events (CMEs).

And unless you build the reactor in space, there's the minor issue of how to get it up to space. The safest real-life rockets still have a roughly 1% failure rate.

As to the OP, none of the stock engines are particularly realistic. The jets are way overpowered in many ways, the rocket engines way underpowered in TWR terms, and there's no LH2 engines. When you deal with game physics, you need to remember the laws of conservation of fun.

[Moar Boosters]

As to the OP, none of the stock engines are particularly realistic. The jets are way overpowered in many ways, the rocket engines way underpowered in TWR terms, and there's no LH2 engines. When you deal with game physics, you need to remember the laws of conservation of fun.

TBH i think all the engines need rebalancing from a gaming perspective. Currently it's too much of a no-brainer to use the nuclear engines for all orbital stages. Just for gameplay reasons it'd be nice if we had engines with ISP's of 500-700, instead of the big leap from 390 to 800.

It's possible they could introduce more powerful engines such as a realistic nerva, but only once career mode has been fully implemented. If an engine is at the end of the tech-tree and costs about a bazillion dollars, then you'd fully expect it to outperform stuff that was unlocked in the first couple of branches. I'm hoping eventually the game will expand beyond a single solar system, so if that ends up happening we're going to need more efficient engines to get there.

[FITorion]

If they add other solar systems the transition to the next system will likely be a loading screen. However in order to get said loading screen some super expensive and efficient rocket engine would probably be required as you say.

[Idobox]

that's an open cycle engine. A closed cycle engine would be heavier, but wouldn't spew radiation all over the place

I feel I have to precise the Nerva didn't spew significant amounts of radioactive waste. The core is solid, and hydrogen flows through it and is ejected. The nuclear fuel is in airtight containers and never touches the hydrogen, so the only dangerous exhaust would be the odd atom teared away from the containment unit, or the even rarer atom of tritium (safe unless you ingest it).

That being said, you still have radiation from a lightly shielded reactor, and a lot of fun cleaning when one crashes.

[vidboi]

...There's the minor issue of how to get it up to space. The safest real-life rockets still have a roughly 1% failure rate.

This is the main reason against using nuclear engines. The proposed nuclear replacement for the S-IVB would require the power output of a large nuclear power station. Imagine what would happen if there was a launch failure: we're talking about a nuclear disaster that makes Chernobyl look small time.

Also, in terms of the game, it's all just a matter of balance. LV-Ns are powerful enough as they are.

[Gilliam]

I actually didnt think of the radiation much. but as far as what I have read, the risk of fallout with NERVA engines was estimated to be really low. NERVA engines were only ever intended by NASA to function as upper stages, firing only once in space. NASA also believed that if a launcher blows up, the engine would remain intact due to it being incredibly tough and heat resistant. They might have actually given the engines their own parashutes.

I can see though, how radiation shielding is a big issue. Its not like you can slap a bunch of layers of lead on there, because thats way too heavy. As for KSP game mechanics though, all the more reason to make the LV-N engines heavier due to radiation shielding requirements, but better in other aspects .

@numerobis: i like your mentioning of the law of conservation of fun, and i guess thats worth a thread on its own. The question really is, what do you find fun in KSP. For me, I really find it fun to optimize, find the limits of how well certain vehicles work. And I dont mind reading a bunch of wikipedia articles to reach that goal. So for me, more depth in the game = more fun .

[RabidMonkey]

In an attempt to bring "real life" nuclear engines to the game, I made some configs for the Timberwind Project from the 1990s. Using the stats from the project, they were totally game-breaking. We're talking single-stage-to-Eeloo levels of overpowered.

[numerobis]

If you buff up the LV-N to be have NERVA specs, but change nothing else, then the LV-N is basically the best engine for all stages all the time. Current engines have a reasonable balance, where no engine is always better for everything (the mainsail has too much thrust for small craft, and the Isp is low; the T30 has lower TWR; the 48-7S requires a large part count, etc). You'd have to adjust all the engines.

There was talk of making the LV-N be politically difficult in KSP also; that was in the distant future where the mythical career mode was implemented. Things have progressed to where there's a career mode, though you still don't have budgets or politicians to placate.

[123nick]

all i want in ksp is a Bimodal nerva, maybe even Trimodal? then i would be very happy

[lazarus1024]

Easy way, introduce a 2.5m NERVA engine that weighs a bunch more and produces around 300 thrust.

The 1.5m sucker is smaller than a real life NERVA engine, so it makes sense that it produces less thrust. It also weighs somewhat less than a real life NERVA engine does.

So a heavier, higher thrust 2.5m NERVA engine makes complete sense.

I am hoping with the revised research tree coming with the next release, engine improvements is one of the things we get.

The stuff in KSP is rather good, but the engines are low in terms of ISP for everything.

A good LH2-LOX engine can produce in the 410-440s range and as mentioned, the better NERVAs in testing before the program was cancelled hit 850s and in theory with some improvements based on 1970's tech, let alone 21st century stuff, probably could have hit over 1,000s, maybe even 1,100s on a solid core design.

So the engines are they are, are fine...but it would be nice if you can unlock something up around tier 5 or 6, that, say, is "Fuel injection optimization - Kerbal scientists realized that if they just inserted part A in to part B backwards, the explody thingies worked so much better!" that increases ISP on all liquid fueled engines by 5% with a tier about that that increases IPS by another 5%.

Have something similar in its own branch for NERVA engines.

Maybe also have something that is materials advancements that reduce the base weight of engines and fuel tanks by 5% or something.

That sort of thing. That way there is more than just parts that you can unlock, you can make parts better. At least slightly.

[hobbsyoyo]

Given that the solar system is 1/10th the size of our actual solar system, I'd say KSP's nuclear engine works great and needs no improvement. If you don't think so, just edit the part config file to match the values you think it should have.

1/10 the size for now....

They aren't done adding planets yet and the ling burn times are only going to get longer. I am all in favor of better nuclear engines. Plus, the NERVA program was built with 60's hardware and materials; if they built it now they could make it withstand heat better which means more efficiency and thrust.

[SRV Ron]

Easy way, introduce a 2.5m NERVA engine that weighs a bunch more and produces around 300 thrust.

Nova Punch has such a engine. Also, they now have two intermediate compact NERVAS. People can play with those and see if they like them or not.