Is there any reason to not use nukes for interplanetary anything?

[maltesh]

LV-N's are harder to design in due to their size and the way the fairings split off

I generally design so they don't have farings. Even with the extra mass, for the vast majority of my vacuum spacecraft and landers (which tend to be in about the ten ton range or higher), two radial LVNs beat a single chemical rocket of any type for delta-V. Having more than twice the specific impulse really is that good.

It will be interesting to see if the upcoming changes that have atmospheric pressure change thrust instead of fuel consumption make me reconsider using LV-Ns on Dunar landers.

[cantab]

It depends on the delta-V you need and the TWR you want.

Delta-V wise, a Duna return trip is about the break-even point, where chemical engines are still competitive. For the Mun and Minmus chemical engines generally beat nuclear on a one-trip basis, while for Dres or the Joolian system nuclear engines generally win.

TWR-wise, if you want a Kerbin TWR of 1 or more then nuclear engines are generally a bad choice.

The other main consideration is overall size. The LV-N is good on most ships, but for very small ones it's too heavy, while for very large ones you'd need to use too many.

But ultimately, do what works for you. For example if you have a tried and tested lifter that puts 40 tons into LKO, then does it really matter whether you use a 40 ton chemical engined ship or a 30 ton nuclear engined one?

[Brainlord Mesomorph]

I have a lot of safety and political constraints for how I employ nukes.

Best,

-Slashy

Given that these are Kerbal "Atomic Motors" that somehow consume oxidizer. and that these are Kerbals, who can survive falls from great heights, don't need food, and have perfected some type of infinite rebreather technology, I figure, what doesn't kill 'em makes 'em stronger.

[wumpus]

One thing I haven't seen mentioned is that you could build a "space tug" (or more likely, just a stage) of plenty of nukes to use just a few periapsis kicking, then send the main vehicle off with less (or even just one) nukes. Once you've built up enough deltaV to be in a highly elliptical orbit between Mun and Minmus (~900m/s), you are nearly there to Eve, Duna (~150,170 respectively). You are "merely" half-way to Jul, Elo, and Moho (and I'm ignoring capture and aero-braking which will count for some huge burns if you are not aerobraking due to "real aero" [mods or >=1.0] or lack of atmosphere (and Dres gets stuffed here due to still needing ~400m/s and no possibility of aerobraking)). In any event, returning a stage that has a perigee already at a parking orbit is trivial, although it might take some time to land specifically on KSC if you wish to do that.

[cerberusti]

It depends, I would say ten tons is about the point where the lv-n becomes dominant.

It is possible to get anywhere without hitting that weight, and to some places with a good amount of payload space when doing so. The most efficient craft which can get to a given planet will not use it. Once you get above that it is pretty much the best engine any time you want more than 5k dv in a stage and do not need a high TWR. Even if you do need a high TWR, strapping on a few 48-7s to boost it during just that part of the mission can help (such as lifting off from a high gravity moon, or trying to slow down enough to obtain orbit around moho.)

[Space Peanut]

People have mentioned this in passing but I have to say that designing ships that can get to orbit with sufficient numbers of LVN's to do the job in decent time is something I really struggle with. Have tried launching the engine blocks seperately and docking to the main ship but often it just ends up a wobbly mess.

[John FX]

As has been said, it depends on the desired TWR.

This is a series of charts started by Tavert which show the best engine for particular payload masses charted against a desired Dv. Each chart shows the results for a particular TWR either in vacuum or atmosphere. When I get time I plan on updating the charts for the current set of engines.

http://forum.kerbalspaceprogram.com/threads/45155-Mass-optimal-engine-type-vs-delta-V-payload-and-min-TWR

[Laie]

People have mentioned this in passing but I have to say that designing ships that can get to orbit with sufficient numbers of LVN's to do the job in decent time is something I really struggle with. Have tried launching the engine blocks seperately and docking to the main ship but often it just ends up a wobbly mess.

Huh? If "decent time" means three minutes rather than fifteen, then Nervas are the wrong tool to begin with.

If your poblem is weight, something is wrong: almost always, nukes save more than their own mass in fuel, so the weight on the pad should be less, not more.

If your problem is structural, some pics would be in order; I'd also suggest you start your own thread for your particular problem.

But groping in the dark without pics, I guess you try to place the LV-Ns in the stack. I suggest you make them as outriggers instead: Just slap "Structural Fuselage"s (look like a jet fuel tank) to the side of a big tank; place Nervas on the fuselages. No fuel lines needed.

Benefits:

• the vessel can be built of relatively few, relatively big tanks (good for stability)
  
• the fuselages provide attachment points for probe cores, reaction wheels and the likes, again moving those out of the main stack (even better for stability)
  
• the Nervas can be run alongside the lower stage's engines (they may be insufficient to do it alone, but they can help a big deal)
  
• no worries about those split engine fairings because there aren't any.
  

Downsides: every nuke increases the part count by two; another 100kg of dead weight per engine.

[HL][/HL]

Edit, because a pic says more than 100 words:

This vessel makes no sense, but it displays the radial nukes.

Edited February 15, 2015 by Laie

[Renegrade]

If your poblem is weight, something is wrong: almost always, nukes save more than their own mass in fuel, so the weight on the pad should be less, not more.

Only if your target TWR is below 1.0 - a craft with a design TWR of 1.0 or higher is definitely carrying more LV-Ns than they're saving in fuel compared to 48-7S or KR-2L designs etc.

(obviously this TWR is unnecessary for many space-y uses, but there is a big quality of life advantage to high-TWR interplanetary stages )

[Sirrobert]

If your poblem is weight, something is wrong: almost always, nukes save more than their own mass in fuel

Unless offcourse you are moving probes, in which case the entire transfer stage + probe masses less than a single NERVA

[Technical Ben]

They are worth it in my opinion. More so if you are using the same ship over and over. Leave it in orbit and then just refuel so you dont need to lift the nuke engines more than once. Personally I never use any other kind of engine in space. Landers and Launchers are different but any long distance moving I do is with Nukes.

This. The simplicity also means 9/10 I use nukes for landers too for moons and anywhere without an atmos. As I can carry 1 or 2 nukes, use for interplanetary, then for landing, over carrying nukes and landers.

Though... should I get a faster PC and more be able to play at a higher FPS with more parts, then I would have no worries on taking a dedicated lander with a small chemical engine.

PS, Laie's picture is similar to some of the really large one time launch fuel tugs I use. So as to avoid refuel in orbit, I send up the tank, engines and launch all in one craft. Either with radial nervas to allow them to assist the launch, or central so as to allow a lower TWR when in orbit (but never as the main lifting force!). On smaller launch craft, they are just stacked on the top stage (sometimes in reverse! ).

Edited February 15, 2015 by Technical Ben

[Raptor9]

Unless offcourse you are moving probes, in which case the entire transfer stage + probe masses less than a single NERVA

Agreed.

There was a NASA study done about comparing logistics required for assembling a Mars Transfer Vehicle with either conventional chemical rockets or nuclear thermal rockets. The paper showed that it would not only require less launches to assemble a NTR-powered vehicle capable of making it to Mars and back, but the chemical rocket version would have to be jettisoning fuel/engine modules left and right to save on mass (due to the lower ISP). The mission profile and payload will always dictate your craft design, but space travel in KSP and in real life is about efficiency. The less efficient your equipment and/or designs are, the more "stuff" you have to take with you, requiring you to put more "stuff" in orbit, requiring more rockets, fuel, parts, money, etc....its a spiral of diminishing returns.

EDIT: Found it: Human Exploration of Mars Design Reference Architecture 5.0 The charts on pages 41 and 43 shows the main differences I was referring to. The chemical option required five additional payload launches to low Earth orbit and an additional 400 metric tons of rocket greatness to accomplish the same mission profile.

Edited February 15, 2015 by Raptor9

[tsotha]

People have mentioned this in passing but I have to say that designing ships that can get to orbit with sufficient numbers of LVN's to do the job in decent time is something I really struggle with. Have tried launching the engine blocks seperately and docking to the main ship but often it just ends up a wobbly mess.

One thing to remember with LV-Ns is you don't have to wait until you reach orbit to use them. By 2000 meters an LV-N's ISP is already 411 (more efficient than a Mainsail in a vacuum), and by 20km it's 789. I usually fire them up as soon as the SRBs are clear, and that little bit of extra thrust over most of your boost can make a big difference in terms of what's practical.

[MarvinKitFox]

Valid point, but aside from funds, does it ever make sense to not use them (im from the era before career even existed, so i still play sandbox all the time).

The only time it does not make sense to use Nukes, for a high delta-v trip, would be if you had even more efficient engines.

Such as the Ion Drive.

As great as the performance increase from Rocket engine to Nuke is, the increase from nuke to Ion drive is four times better still.

(The pains too: Price. Minimal TWR. Difficulty of build)

[eddiew]

There's definitely a minimum size where having the weight of a nuke onboard does not improve matters over leaving it off entirely. Shows up most in small SSTOs, because you'll have to be carrying other engines already.

As an example, this is about the smallest thing I've found to benefit from an LV-N. The nerva is 1/6th of its total mass and has about 3t of fuel left to work with by LKO. If it had only 2t of fuel, burning it 2.2x as efficiently as the rapiers that are already on board, then you've carried a 2.5t engine to save 2.4t of fuel - not an effective choice

Edit; the screenshots are for a Munlanding, but this can of course go orbit Eve or Duna and return as an alternative, whence comes the interplanetary bit

And of course really tiny probes are better with ion engines, and can be lifted to orbit in an SSTO for maximum efficiency

Edited February 16, 2015 by eddiew

[Acet]

As others pointed out, in career mode, anything with lowish Dv requirements and low weight is much cheaper done with the lighter engines, either the 909 or the 48-7S.

I've done the entire Kerbal system exploration with those engines, both on probes and capsules. Furthermore, probe exploration of the Eve and Duna areas (sending a ship with two probes to each planet) was done on the cheap using that tech for the interplanetary transit and then IONs on each probe for the final adjustments around the planets and the moons. While nuclear engines would've given me some peace of mind (not much Dv left in the 48-7S-powered interplanetary transit ship after aerobreaking), they would've meant almost doubling the weight of the whole thing and going up one or two tiers of launcher instead of the "small pile of entry level boosters plus one 909" that I used. A cost of 70k+ instead of 30k.

In fact, like somebody else in this thread I have a very cheap (17k, if I remember correctly) design for a Kerbol system probe which, with exception of the ION engine on the probe, only uses low-tech propulsion (2x RT-10, 909 + FL-T400)

In my previous run through the game I was using nukes all over the place as well as asparagus staging in the craziest of configurations. This time around I'm discovering just how much you can do with little and on the cheap (boosters finely tuned to just barely hit terminal velocity, small engines, reusable air-breathing VTOL launchers, probes, multi-part missions assembled in space, separate refuelling missions) and it's quite surprising just quite small the minimum configuration for exploration is and how much you can do with reusable launcher cores surrounded by multi-staged rings of boosters all tuned to not exceed terminal velocity.

That said, anything bigger than probes going beyond the Kerbal moons does require nuclear engines and a 3 kerbal capsule is maybe the maximum one can send to the Mun or Minmus without a nuclear engine. Even my current mission to send 3 probes at once to explore Jool and its moons (with enough Dv in each that they can hopefully cover 2 moons each) did end up requiring nukes.

Also, as soon as we are talking about heavier and/or reusable vehicles (space stations in Duna or beyond orbit, mining ships), then nukes are the only reasonable choice.

[John FX]

OK, just to show that the NERVA is not the best at all situation, here is the chart of the best engine to use if you do not limit the number of engines on a craft for chosen payloads and Dv required for a TWR =1

As you can see, the NERVA is not the best engine for ANY payload or desired Dv...

This is the chart for TWR=1 and a single engine, just for comparison

As you can see, if you require a decent TWR the NERVA is not good at much...

Edited February 16, 2015 by John FX

[Raptor9]

IMHO, a good TWR isn't going to do much good if you run out of fuel before you get to where you're going. While TWR should be a consideration, I believe it's secondary to ISP in regards to interplanetary transfer operations. Useful charts though.

[FishInferno]

Nukes are good for "there and back" trips because you need to pack less fuel, but for one-way trips just throw a chemical engine on there for a higher TWR because you won't be coming back anyway.

[Sirrobert]

IMHO, a good TWR isn't going to do much good if you run out of fuel before you get to where you're going. While TWR should be a consideration, I believe it's secondary to ISP in regards to interplanetary transfer operations. Useful charts though.

And ISP is useless if you have to burn for 30 minutes at a time to get anywhere

[Brainlord Mesomorph]

And ISP is useless if you have to burn for 30 minutes at a time to get anywhere

Not useless at all!

I have burns of an hour or more. (See the tutorial in my sig,)

All I need is enough thrust to blow out a birthday candle and enough fuel to do it for an hour and a half!

[Renegrade]

That said, anything bigger than probes going beyond the Kerbal moons does require nuclear engines and a 3 kerbal capsule is maybe the maximum one can send to the Mun or Minmus without a nuclear engine. Even my current mission to send 3 probes at once to explore Jool and its moons (with enough Dv in each that they can hopefully cover 2 moons each) did end up requiring nukes.

My last big BTSM playthough ended with the nuke engine node being the only one unbought. I pretty much went everywhere in the Kerbol system on chemical engines, including beyond solar power range (BTSM solar cuts out entirely around Dres's AP) with dual 8-ton reactors (the equivalent mass of four Mk1-2s).

Nukes are not required at all. In a lot of those situations, they can help you reduce mass (and TWR ), but you never NEED them.

(Heck a BTSM Mun mission is something like the mass of an Apollo CSM once it's in space, done with nerfed versions of the old pre-ARM chemical engines)

I have burns of an hour or more. (See the tutorial in my sig,)

:C

Your tutorial lied to me! It said "no math!" "D = (Ejection Angle -90) / 3.38"?? Math! ARGH! *digs an anti-math bunker and hides in it*

(also see my comment above about 'quality of life'. PE kicking isn't quality of life dammit >.<)

[John FX]

IMHO, a good TWR isn't going to do much good if you run out of fuel before you get to where you're going. While TWR should be a consideration, I believe it's secondary to ISP in regards to interplanetary transfer operations. Useful charts though.

If that's the case then there is still only a limited set of parameters where the NERVA is the best engine

It's somewhere between TWR=0.4 with a heavy load

and 0.9 with a lighter load

The NERVA is only best at ultimate Dv when your TWR is between 0.4 and 0.8 which is a reasonable TWR range to be honest.

If, as you say, ISP is king over TWR then this chart is the one you should be taking your information from.

The TWR=0.1 chart.

This shows that the best engine for most situations, if you are not concerned with TWR or part count is the ION thruster.

[Technical Ben]

"The NERVA is only best at ultimate Dv when your TWR is between 0.4 and 0.8 which is a reasonable TWR range to be honest."

This. Are you telling me I get more DV for interstellar transfers with any other engine? If I do, I'll redesign my fleet!

If not, then I'll still not spam the 48-7S just yet. Though a Skipper powered fleet might increase the size of my craft 2fold!

PS, can you redo that graph without the ion, as it's of cause the best for probes and/or long burns, but not really optional for any manned or large craft (due to part lag of having to install 100s of ions ).

[Renegrade]

This shows that the best engine for most situations, if you are not concerned with TWR or part count is the ION thruster.

Does that take into account the necessary mass for power generation? Although I guess that would be a relative thing - you could kick the PE for sixty or seventy orbits using a single ox-stat and a Z-100. (QUALITY OF LIFE ISSUE)

(and until Squad fixes #lolmassless, you can spam oxstats and radial batteries, although I consider that an exploit, plus my Horrible Nerf ends that violently, plus the #loloxstat garbage)

This. Are you telling me I get more DV for interstellar transfers with any other engine? If I do, I'll redesign my fleet!

If you're making a vessel with a TWR of 1.0 or higher, most other engines will give you better delta-v.

PS, can you redo that graph without the ion, as it's of cause the best for probes and/or long burns, but not really optional for any manned or large craft (due to part lag of having to install 100s of ions ).

I did some ion Mun landers that had like .. 1.6 or so Mun TWR (that's just a bit under 0.3 Kerbin TWR) - they only needed two engines or so.

At least...they needed only a handful of engines before my Horrible Nerf put the ion back to 0.5 kn thrust~ muahaha~