Most efficient combo of Poodle and Nukes for a Big Orange Tank O Gas?

[dire]

Hi all, I'm wondering about this situation:

 

 

What happened is, I had a traveling science / tourist bus that went to Minmus and the Mun and then ran out of gas. So I sent a rescue tanker full of fuel to tug it to its eventual destiny of a fiery re-entry through Kerbin's upper atmosphere.

 

My question is, assume the tugboat had started its voyage to the Mun with exactly one full tank of Big Orange Liquid Fuel and Oxidizer, as well as the empty LF tanks and the empty adapter tank seen on the drone (which were spent to circularize into LKO. What is the most fuel-efficient combination of Poodle and Nuke to reach the Mun? I ask because it occurs to me that I'm currently lugging probably 10 or 15 tons of oxidizer around with no liquid fuel I can burn with it. Initially I thought, well, the nukes have ISP 800, they are twice as efficient even if I have to haul a bunch of oxidizer around afterwards. But then I noticed my available delta-V was dropping faster than Flight Engineer thought it should, even over the course of a single burn. Now, I'm not the world's best pilot, but even I seldom spend 600 delta-V on a 400-DV munar exit burn.

Ideally I'd like to dock the Munpocalypse to my LKO station to transfer all kinds of science readings into the big lab there. However, I'm not sure I have enough delta V to make interception and I'm safer just de-orbiting the Munpocalypse and discarding the redundant science. What do you guys think?

[Tonka Crash]

I'd get a mod to let you change the fuel tank type to on liquid fuel only. There are several out there, but I like this one:

 

[dire]

1 hour ago, Tonka Crash said:

I'd get a mod to let you change the fuel tank type to on liquid fuel only. There are several out there, but I like this one:

 

On a well-planned mission I certainly could, but in fact your solution would require one additional suborbital stage on the rocket, which already has enough fuel to complete the mission, because I need the TWR of the poodle to even get into orbit.

To clarify, I already have a mod that lets me switch fuels in the hangar. I already calculated out my delta-V in the hangar, added dedicated liquid fuel containers and so on. A solution that starts in the hangar doesn't answer my question. Further, an additional reason I want to keep the poodle aside from TWR is that one of my mods changes the way nuclear engines work so they throttle slowly up and down from their maximum output, which makes precision maneuvers (such as an intercept to an erratic lunar orbit) either tedius, difficult if not impossible, or both. The poodle allows me to turn off my high-efficiency, low-precision, low TWR nuclear engines and use a high-precision, high-TWR engine that is instant-on, instant-off.

While I could use monoprop, I actually had another mission parameter that the total entity must have at least 1000 monoprop, so although I brought extra I didn't want to rely on it for maneuver nodes (also it turns out the monoprop setup barely squeaks out 0.5 m/s of delta-v per second, so I don't want to sit holding down the 'H' key for ten minutes at a stretch).

 

Edited November 15, 2018 by dire

[Pecan]

The most fuel-efficient is one nuke.
Anything more adds mass but contributes less, ergo reduces fuel-efficiency.

If you *must* keep another engine then use the lightest, highest ISP available.  That'd be the terrier

Edited November 15, 2018 by Pecan

[DerekL1963]

Honestly...  I'm not even sure I understand what you're asking.   It makes no sense to combine Poodles and nukes, and how much oxidizer you have at the Mun is...  not at all relevant if you're launching a rescue mission from Kerbin.  (I think that's what you're asking.)
 

Not to mention, there's no single "most efficient" combination...  It's going to vary with the weight of your vehicle, how long a burn you can tolerate, etc... etc...
 

11 hours ago, dire said:

A solution that starts in the hangar doesn't answer my question.

All solutions ultimately start in the hangar.   In this scenario, it's particularly important because you need to substantially increase the ratio of fuel:oxidizer tankage to avoid lugging around oxidizer that you won't be using.
 

[dire]

1 hour ago, DerekL1963 said:

Honestly...  I'm not even sure I understand what you're asking.   It makes no sense to combine Poodles and nukes, and how much oxidizer you have at the Mun is...  not at all relevant if you're launching a rescue mission from Kerbin.  (I think that's what you're asking.)
 

Not to mention, there's no single "most efficient" combination...  It's going to vary with the weight of your vehicle, how long a burn you can tolerate, etc... etc...
 

All solutions ultimately start in the hangar.   In this scenario, it's particularly important because you need to substantially increase the ratio of fuel:oxidizer tankage to avoid lugging around oxidizer that you won't be using.
 

I was asking, given the station in my first post, how do I best get a result like the pic above? (They made it home safe on a polar orbit. The big orange later detached and circularized to become a polar refueling station, while the hab station de-orbited safely, minus the big yellow monoprop tank exploding halfway down and the other 2.5m parts crumpling on impact).

 

1 hour ago, Pecan said:

The most fuel-efficient is one nuke.
Anything more adds mass but contributes less, ergo reduces fuel-efficiency.

If you *must* keep another engine then use the lightest, highest ISP available.  That'd be the terrier

 

Not sure how you are getting -into- orbit on a Terrier and a nuke, not to mention sitting patiently waiting for a single nuke to impart 600 dv on a 60 ton ship.

Edited November 15, 2018 by dire

[DerekL1963]

1 minute ago, dire said:

Not sure how you are getting -into- orbit on a Terrier and a nuke, not to mention sitting patiently waiting for a single nuke to impart 600 dv on a 60 ton ship.

o.0  I didn't say anything about a Terrier and a nuke...  As far as waiting out long burns, I have a computer that multitasks.  I'm currently on the forums while MJ drives a rover across Duna.

 

13 hours ago, dire said:

also it turns out the monoprop setup barely squeaks out 0.5 m/s of delta-v per second, so I don't want to sit holding down the 'H' key for ten minutes at a stretch

That's what the "fore by throttle" setting on your RCS blocks are for.  Turn that setting on, firewall the throttle, turn your RCS on, and you're set.

[Pecan]

2 hours ago, dire said:

Not sure how you are getting -into- orbit on a Terrier and a nuke, not to mention sitting patiently waiting for a single nuke to impart 600 dv on a 60 ton ship.

I wouldn't dream of launching with a nuke and terrier but then I also wouldn't dream of taking my launch vehicle to Mun.

You didn't ask how to launch but what the most fuel-efficient engine arrangement was.  It's one nuke.  If you *also* introduce the requirement for a LFO engine, it's one nuke and a terrier.  If you *also* insist the LFO engine is a poodle, it's one nuke and one poodle - there is no escape from the fact that including lower-ISP engines, or more engines always reduces the fuel-efficiency, because you're adding mass.

Given that any LFO engine is going to consume fuel and oxidiser at exactly the same, 9:11, ratio that they are stored in fuel tanks any fuel required for nukes will always be extra.  No possible arrangement will even-out the discrepancy between LF and O consumption, unless you abandon the nukes altogether.  In that case your most fuel-efficient engine arrangement is one terrier.

[sturmhauke]

This seems to be what we in the software world call an XY problem. Your actual problem X, which you haven't directly stated, seems to be "How do I get a science/tourist craft back to LKO from the Mun?". But what you asked was how to complete your partial solution Y, "What is the most efficient combo of engines to use on the rescue craft?"

Fuel efficiency is not really the most important concern for a rescue craft. High thrust and maneuverability are more important - thrust because it will have additional mass to contend with after recovering the target, and maneuverability to handle rendezvous and docking.

To that end I'd say forget the Nervs. You're only going to the Mun, so you're not saving that much fuel anyway, and they're heavy. The Poodles are probably a good compromise - good vacuum isp and decent TWR. Don't forget your heavy RCS systems, and a Klaw if your target vessel lacks docking ports.

[dire]

22 minutes ago, sturmhauke said:

This seems to be what we in the software world call an XY problem. Your actual problem X, which you haven't directly stated, seems to be "How do I get a science/tourist craft back to LKO from the Mun?". But what you asked was how to complete your partial solution Y, "What is the most efficient combo of engines to use on the rescue craft?"

Fuel efficiency is not really the most important concern for a rescue craft. High thrust and maneuverability are more important - thrust because it will have additional mass to contend with after recovering the target, and maneuverability to handle rendezvous and docking.

To that end I'd say forget the Nervs. You're only going to the Mun, so you're not saving that much fuel anyway, and they're heavy. The Poodles are probably a good compromise - good vacuum isp and decent TWR. Don't forget your heavy RCS systems, and a Klaw if your target vessel lacks docking ports.

I would argue that this is a case of "Did not read the manual," and therefore "Scope creep". I didn't actually expect anyone to read my summary ("My question is, assume the tugboat had started its voyage to the Mun with exactly one full tank of Big Orange Liquid Fuel and Oxidizer, as well as the empty LF tanks and the empty adapter tank seen on the drone (which were spent to circularize into LKO. What is the most fuel-efficient combination of Poodle and Nuke to reach the Mun? ") but I was at least hoping people would look at the pretty pictures.

I should also note that the tug strikes a happy medium between your solution ("Ditch the nukes, use just the poodle") and the previous solution ("Ditch the poodle and one nuke, use just one nuke for the tug"). I already made that design decision. The tug is in space. 

I am thinking about re-using this design and sticking a pair of these tugs to drive a large interplanetary craft (four kerbals, two years of hab and food is probably the minimum I can do for a manned mission to Duna). I'm curious about the best way to use a big orange tank of fuel with them.

Edited November 15, 2018 by dire

[Zhetaan]

On 11/14/2018 at 8:14 PM, dire said:

My question is, assume the tugboat had started its voyage to the Mun with exactly one full tank of Big Orange Liquid Fuel and Oxidizer, as well as the empty LF tanks and the empty adapter tank seen on the drone (which were spent to circularize into LKO. What is the most fuel-efficient combination of Poodle and Nuke to reach the Mun?

One Poodle.  If you insist on taking the O in LFO, then nukes have to haul that as dead weight.  That makes it payload, and that utterly destroys your delta-V because it not only eliminates fuel from the wet mass side of the mass ratio, but it also puts that former fuel on the dry mass side.  You cannot win on efficiency by using a nuke with a full bipropellant tank.

Another thing to remember is that the mass ratio of the rocket equation doesn't care what the fuel is, but in order to compare engine efficiency (in the form of specific impulse) as an independent variable, you need equivalent masses of fuel (and rocket).  Putting aside the actual masses involved, let's just say that the big orange LFO tank has twenty total mass units of fuel, in the proportion of nine LF to eleven O.  The Poodle, which consumes LFO in that proportion, thus has the use of all twenty mass units.  A nuke, to be fairly compared, also needs twenty mass units--but those units need to be all LF because that is what the nuke uses as fuel.  Failure to provide that results in a rocket that has less than half the fuel load of an equivalent LFO engine, and to measure the nuke's performance against an LFO engine that has not been so hobbled is not a fair comparison.

To wit: your example rocket is 53.589 tonnes from the readout.  Subtracting the 2.7436 tonnes of LF and 12.3794 tonnes of O that you have in the image gives a dry mass of 38.466 tonnes.  Assuming that the orange tank is full adds 32 tonnes of fuel mass for a total of 70.466 tonnes of rocket.

The Poodle has a vacuum specific impulse of 350, so the rocket equation for this vessel is:  delta-V = 350 * 9.80665 * ln (70.466 / 38.466), which gives a total delta-V of 2077.78 m/s.

The Nerv has a specific impulse of 800, but it can only use the 14.4 tonnes of LF in the tank; the remaining 17.6 tonnes of O must be accounted as dry mass.  The total mass remains the same, of course, but the dry mass for this rocket under Nerv thrust becomes 56.066 tonnes.  The equation for this is delta-V = 800 * 9.80665 * ln (70.466 / 56.066), which gives a total delta-V of 1793.45 m/s.

Shut the nuke off and leave it off.  It's still three tonnes of dead weight, but that's better than 17.6 tonnes of useless oxidiser.

[Plusck]

20 hours ago, dire said:

I would argue that this is a case of "Did not read the manual," and therefore "Scope creep". I didn't actually expect anyone to read my summary ("My question is, assume the tugboat had started its voyage to the Mun with exactly one full tank of Big Orange Liquid Fuel and Oxidizer, as well as the empty LF tanks and the empty adapter tank seen on the drone (which were spent to circularize into LKO. What is the most fuel-efficient combination of Poodle and Nuke to reach the Mun? ") but I was at least hoping people would look at the pretty pictures.

I should also note that the tug strikes a happy medium between your solution ("Ditch the nukes, use just the poodle") and the previous solution ("Ditch the poodle and one nuke, use just one nuke for the tug"). I already made that design decision. The tug is in space. 

I am thinking about re-using this design and sticking a pair of these tugs to drive a large interplanetary craft (four kerbals, two years of hab and food is probably the minimum I can do for a manned mission to Duna). I'm curious about the best way to use a big orange tank of fuel with them.

The problem - as I see it - is that your question is essentially unanswerable.

There is no possible "efficient" solution if the starting position includes both (1) a 9:11 LFO mix, and (2) Nukes (any number).

Each engine has a "peak efficiency" curve. Or rather, several such curves depending on the variables that you are measuring efficiency by. The variables you can use are firstly payload, TWR, dV. Then secondly cost, fuel mass, playing time, required additional equipment (i.e. part count), elegance. From the first three (payload, TWR, dV) pick one as a fixed variable, and then you can relatively easily plot the other two as a curve.
And good example of this is here: https://meithan.net/KSP/engines/
It's a bit old now, and some engines might be wrong, but the principle is sound. This is the graph for a 40 tonne payload:

Each "step" in the graph represents an additional engine. So you can see that for this 40t payload, if you want more than 0.5g thrust, nukes are only the best option if you have 7 or 8 of them. And you intend to run them for a dv of about 1800 - 2500 m/s.

So the trouble with nukes is that they are so heavy that they need to keep burning for quite a while before their efficiency makes up for their own mass. Or the payload has to be so big that the mass of the engine becomes less relevant.
And in all cases, if you start looking at the other "efficiency" variables, such as cost, manouverability, low part count and so on, there will always be a better option, and that option is often the Poodle... And if you have to start calculating how much extra liquid fuel you need to carry to balance the LFO mix burnt by one of the chemical rocket engines, you lose that last possible efficiency benefit: your playing time.
The pic above doesn't show you the next best option, but you can see that if you follow the link and use Meithan's tool. In most of the usual scenarios, that second-best option is either the Aerospike or the Poodle. One of those two has a distinct advantange for cost and manoeuverability (and therefore part-count); no guesses which...

So, no, there is no real answer to your question. If you want to move a big orange tank somewhere, you definitely don't want to use a nuke to do it. If you have nukes onboard already, you'd be best ramming them into something and destroying them so you can lighten the load.

However, creating a nuke tug is definitely a good idea. Create a re-usable, drag-and-drop nuke-based interplanetary booster, with radiators, drop tanks, torque wheels and suchlike, and you will definitely end up with a very efficient system, certainly better than any LFO alternative (at least until the Rhino becomes a reasonable option, for huge ships). But don't try to mix it with any LFO engine, you'll waste either the benefits, or your time, or both.

[sturmhauke]

1 hour ago, Plusck said:

... https://meithan.net/KSP/engines/
... This is the graph for a 40 tonne payload:

Ooh, that is cool. I might have to play with that code and see if I can do an updated version.

[dire]

2 hours ago, Plusck said:

The problem - as I see it - is that your question is essentially unanswerable.

There is no possible "efficient" solution if the starting position includes both (1) a 9:11 LFO mix, and (2) Nukes (any number).

Each engine has a "peak efficiency" curve. Or rather, several such curves depending on the variables that you are measuring efficiency by. The variables you can use are firstly payload, TWR, dV. Then secondly cost, fuel mass, playing time, required additional equipment (i.e. part count), elegance. From the first three (payload, TWR, dV) pick one as a fixed variable, and then you can relatively easily plot the other two as a curve.
And good example of this is here: https://meithan.net/KSP/engines/
It's a bit old now, and some engines might be wrong, but the principle is sound. This is the graph for a 40 tonne payload:

Each "step" in the graph represents an additional engine. So you can see that for this 40t payload, if you want more than 0.5g thrust, nukes are only the best option if you have 7 or 8 of them. And you intend to run them for a dv of about 1800 - 2500 m/s.

So the trouble with nukes is that they are so heavy that they need to keep burning for quite a while before their efficiency makes up for their own mass. Or the payload has to be so big that the mass of the engine becomes less relevant.
And in all cases, if you start looking at the other "efficiency" variables, such as cost, manouverability, low part count and so on, there will always be a better option, and that option is often the Poodle... And if you have to start calculating how much extra liquid fuel you need to carry to balance the LFO mix burnt by one of the chemical rocket engines, you lose that last possible efficiency benefit: your playing time.
The pic above doesn't show you the next best option, but you can see that if you follow the link and use Meithan's tool. In most of the usual scenarios, that second-best option is either the Aerospike or the Poodle. One of those two has a distinct advantange for cost and manoeuverability (and therefore part-count); no guesses which...

 So, no, there is no real answer to your question. If you want to move a big orange tank somewhere, you definitely don't want to use a nuke to do it. If you have nukes onboard already, you'd be best ramming them into something and destroying them so you can lighten the load.

 However, creating a nuke tug is definitely a good idea. Create a re-usable, drag-and-drop nuke-based interplanetary booster, with radiators, drop tanks, torque wheels and suchlike, and you will definitely end up with a very efficient system, certainly better than any LFO alternative (at least until the Rhino becomes a reasonable option, for huge ships). But don't try to mix it with any LFO engine, you'll waste either the benefits, or your time, or both.

Thank you, this is an excellent answer. I really appreciate it!