Save my Neidon mission

[bigcalm]

The Situation

3 Kerbals are on board the "Isaac Newton" on a nine year mission out to Neidon (currently 300 days out from Kerbin).  They will run out of food after around seven years.  I am using USI mod with life support, and need to get some additional fertiliser to the vessel before it reaches its destination.  Note that there's a maximum carrying capacity of fertiliser on the Isaac Newton of around seven years worth of fertiliser (==food).

The Options

The "Isaac Newton" is not the only vessel on its way to Neidon.  In the same launch window, I sent two additional vessels that have fertiliser (there are more in the flotilla, but none of the others will help):

Nissee Orbit Spares -- has enough fertiliser to top up, and also has extra docking ports should they be needed.  Slight issue here is that it was intended to serve as the return vehicle, though I can send another return vehicle in the next window.

NIssee Base -- has enough fertiliser to top up, is probably the closest in terms of orbital parameters to the "Isaac Newton" currently.  However, this really needs to make it as it will serve as new fertiliser generation.

Neither have much spare delta-v, though I'm willing to sacrifice either to save the Kerbals.

The Kerbin Launch window for Nissee comes up in around 130 days.

I have correction burns already planned in for all vessels in the flotilla.

Suggestions

I think my plan has to be to attempt to rendezvous the orbit-spares with Isaac Newton some time around year seven, dock it, top the fertiliser up, and then at least the Kerbals will make it to Neidon, and hopefully the base will then be operational and I can generate more fertiliser.  What kind of burns (pro/radial/normal) do I need to make to do this kind of rendezvous where both craft are in roughly the same orbit but not needing to meet for another 5 years or so?  Assuming I burn too much fuel, the used-up vessel won't make Neidon orbit, but that's probably ok - I can send out a replacement from Kerbin to arrive a year later.

The alternative may be the transfer window that's coming up in 130 days or so - I can launch some fertiliser and assuming I keep it to just the fertiliser plus large shunt, I probably have a large amount of delta-v to play with.  Is it possible I could catch up? (with Isaac Newton basically departing the year before).  

 

Edited July 19, 2018 by bigcalm

[Snark]

18 minutes ago, bigcalm said:

3 Kerbals are on board the "Isaac Newton" on a nine year mission out to Neidon (currently 300 days out from Kerbin).  They will run out of food after around seven years.  I am using USI mod with life support, and need to get some additional fertiliser to the vessel before it reaches its destination.  Note that there's a maximum carrying capacity of fertiliser on the Isaac Newton of around seven years worth of fertiliser (==food).

There are a lot of ways to slice it-- hey, this is KSP.   

One of the nice things about a long, slow mission such as you describe is that there's a lot of flexibility in "how much time does the trip take" if you have a craft with a bit of extra dV to spare.

Also bear in mind that there are different kinds of launch windows.  Simple Hohman's not the only option-- you can, for example, launch a craft solar-retrograde so that it falls down to a solar Pe significantly lower than Kerbin's, and then do a large burn at Pe to go zooming out to the outer solar system.

So... why not just launch one more ship, a supply ship?  Make it simply a fertilizer hauler.  I've never used USI life support, so I'm not sure how much mass we're talking about... but if you build a ship that's basically just a fertilizer container and a fuel tank, how big does it need to be?  We're not talking about something really huge, are we?

Assuming it's not too giant, just build a special-purpose fertilizer-resupply ship and launch it to catch up with your crewed ship shortly before year 7.  The dV to get the intercept should be reasonably modest.  You'll likely need a significant chunk of dV to match velocities when you do intercept, but we're not talking something seriously insane, here-- I expect a reasonably designed LV-N ship should be able to manage it.

Any reason not to try that approach?  Simple, direct, shouldn't be too expensive, doesn't require sacrificing or compromising any of the existing ships or mission goals.

[bigcalm]

Thanks Snark.

I think the main problem is that whilst I'm familiar with the standard set of maneuevers - Hohmann Transfer, Plane matching, Circularising, this one is a bit new to me - sending something to rendezvous with a craft that's already on a highly eccentric orbit.  I will try what you suggest though!  Is there a way to figure out the best transfer window from Kerbin? 

Yes, the mass of the emergency fertiliser will be relatively low - roughly 6 tons of payload once I add docking capabilities, and my standard shunt is pretty darn big so I will have a lot of potential dV to play with.

Edited July 20, 2018 by bigcalm

[bewing]

1 hour ago, bigcalm said:

I think the main problem is that whilst I'm familiar with the standard set of maneuevers - Hohmann Transfer, Plane matching, Circularising, this one is a bit new to me - sending something to rendezvous with a craft that's already on a highly eccentric orbit. 

Normal low-speed rendezvous maneuvers can be generalized and used as high-speed rendezvous maneuvers. You go into Target mode on your navball -- you go full throttle and pull your prograde marker on top/in front of your target marker. When you get close you turn around retrograde, burn like hell and push your retrograde marker on top of your anti-target marker. It's not really something where you are looking for a window. You just want to get pointed in the right direction and then go 1 km/s faster than you need to for a hohmann transfer.

Edited July 20, 2018 by bewing

[Zhetaan]

@bigcalm:

You definitely have an interesting problem.  As @bewing said, you're not really looking for a window.  If your two vessels share an orbit, there technically is no window, because they will always have the same relative orbital alignment.  Thus there's no time that is any more or less favourable to make the transfer.  Since you're obviously running Outer Planets Mod, I invite you to look at Polta and Priax; Priax is a simulated Trojan body, so it has no transfer window to Polta (and vise versa) because it never gets any closer or farther away.  The correct solution for this is to burn for an intercept in much the same fashion as you would an asteroid rendezvous.  If you haven't done that yet, then look at the video tutorials (don't try to run one on your own copy of KSP--the tutorial scenarios won't play well with the mods).

To go along with what @Snark said, what you normally think of as a transfer window is actually a window for a Hohmann transfer.  If you're using a different kind of manoeuvre, then different windows apply.  You'll have to determine what is best for yourself, though; I don't know the type of manoeuvre you're planning to use.  Note that if you want to send a second flotilla (or only a single resupply vessel) on a Hohmann transfer, the Hohmann transfer windows for the outer planets all occur about once per year.  This is because the outer planets revolve about the Sun so slowly compared to Kerbin, and over so much more distance on their orbits, that they are relatively almost motionless compared to Kerbin.  That means that you can plan on sending resupply a tiny bit less than once per year if minimum-fuel rendezvous is important to you.

Were I in your position, I would choose to try to rendezvous the Nissee Orbit Spares first.  My reasoning is pragmatic:  the vessel is already near to rendezvous anyway, and it's the more expendable of the two:  whether or not you manage to effect a rescue with Nissee Orbit Spares, you will need the Nissee Base.  If you have that, you can hold out until you run out of habitation (or indefinitely if you aren't using that part of USI-LS).  Furthermore, it is possible for you to lose Nissee Orbit Spares, either physically or functionally, and still have a return vehicle:  you can use the Isaac Newton.  You may have to scramble at Kerbin to catch it, but you should be able to return.  On the other hand, if you lose Nissee Base, you won't be able to get back.  I'm going to guess that you planned the same amount of food and supplies for the return trip as the outbound one, which means that you're still going to need some kind of resupply, so if you lose Nissee Base to a bad manoeuvre, you won't have enough life support either to get home or to make a stand out there.

[Snark]

8 hours ago, bigcalm said:

I think the main problem is that whilst I'm familiar with the standard set of maneuevers - Hohmann Transfer, Plane matching, Circularising, this one is a bit new to me - sending something to rendezvous with a craft that's already on a highly eccentric orbit.  I will try what you suggest though!  Is there a way to figure out the best transfer window from Kerbin? 

Yes, the mass of the emergency fertiliser will be relatively low - roughly 6 tons of payload once I add docking capabilities, and my standard shunt is pretty darn big so I will have a lot of potential dV to play with.

Great, that narrows the scope of the problem.    So, this isn't an engineering problem-- it's a navigation problem.

So, basically you need to figure the following:

• When to launch
• What path to take (i.e. what sequence of burns)
• How much dV it will need (so you can design your ship)

Here's the executive summary of how to pull this off:

1. Eyeball where your target ship will be at intercept time.
   • This is so you know where to aim.
   • It's easy & quick.
2. Figure out which launch window will take you there.
   • This is so you know when to launch.
   • Reasonably straightforward.
3. Figure out what path you'll take, i.e. what sequence of burns.
   • This is so you'll know how much dV you'll need, which is necessary in order to design your rescue ship.
   • This step is the tricky bit, because it's where you figure out most of the navigation. 
4. Design and launch the ship.
   • This one's an engineering problem, not a navigational one.  So I'll assume you've got this covered.
5. Plot the course.
   • This one's just a verbatim replay of the figuring-out that you'll do in step 3 above, so therefore it's easy.

I'll talk you through steps 1 and 2 here, then pause for a breather to see what you think before going on.

 

Step 1:  How to eyeball where the target will be at time of intercept

Well, that's hard to do precisely, but fortunately you don't need to.  For purposes of this step, you only need to eyeball it approximately, and in your particular case that's pretty easy to do. 

Why is it easy in this case?  Because you've got that nice, long seven-years-out-of-a-nine-year flight ahead of your ship.

It means that if you open your map view, you can already see where your troubled ship is going to hit its solar Ap, nine years from now.  So you just need to visually estimate "where is the ship going to be, around 2-3 years before Ap".  And that's easy to do, because those last 2-3 years are going to be when the ship is moving super slowly (because it's waaaaaay out in the outer solar system at the outer end of a very eccentric orbit), so it won't cover much distance during that time.

So all you need to do is look at where your target ship's solar Ap is, and then mentally estimate a point along the orbit that's "a bit" before that.  It helps that you don't have to get it exactly, because as long as you can get it somewhat approximately, that's good enough for launching purposes.  Even though you won't nail it perfectly, the error will be easily corrected by adjusting burns at launch time and won't cost an unreasonable amount of dV.

In other words:  If you picture the solar system as a clock dial with the sun at the center, all you're trying to guesstimate is "at which o'clock will the intercept happen", and as long as you get it within one or two clock-hours, you'll be fine.

(Note:  If this kind of wild-ass guesstimate bugs you and you'd prefer a more accurate picture, you could always just do a quicksave and then timewarp ahead 6-7 years to see where the ship is, then go back and reload.)

 

Step 2:  How to pick a launch window

This also is pretty straightforward, because you have a lot of flexibility here.  A digression into why it's so flexible and why you don't have to worry too much about this, in spoiler section.

Spoiler

Launch windows aren't a hard-and-fast requirement, nor are they a simple binary on/off thing.

Rather, there are certain times that are "ideal" for making the transfer (i.e. require minimum dV)... but it's not as though you have to be right at that moment and if you go outside it there's suddenly a huge stair-step increase in the dV required.  Rather, if you launch a little bit outside the "ideal window", it just needs a little bit of extra dV.  Launching further outside requires more dV, and so forth.

Technically, you can launch any time you like.  It's just a question of how much dV you're willing to spend.  As Kerbin circles the sun, a few points in that orbit will be really good "windows"; and there will be some points that are absolutely terrible; and there will be a really broad swath of points in between that are kinda mediocre.

When dV is very tight-- e.g. you're lugging a heavy payload, so it's an engineering challenge to pack lots of dV into it-- then you kinda need to aim for the window, because you don't have the dV to spare to take a less-efficient trajectory.  But the more dV you have, the less you have to care.  If you've  got 2000 or 3000 m/s of "extra" dV, that can make a huge difference in flexibility of launching.

In this case, you're shipping a pretty small payload (six tons), so it's not too hard to engineer your ship to give it a metric crapload of dV to play with, which gives you a lot of flexibility.  You don't need to worry about "get a really good launch window" at all.  Rather, all you need to do is avoid getting a really terrible one, and that's pretty easy. 

Okay.  So, picking a general launch window is pretty straightforward here because you know what direction you're going (having estimated the intercept location in step 1 above).  You know that you want to eject from the inner solar system going in a certain direction.  So all you need to do is pick where to eject from Kerbin in order to make that happen.

There are a few ways to slice this, but there are two main types of path you could follow, both of which probably involve a similar amount of dV, so you should just pick whichever one is available soonest.  (Yes, there will be some dV difference between the two, but it won't be a huge difference and you've got a fair amount of dV to play with, and time's a-ticking so sooner is probably better.)

Broadly speaking, the two choices of trajectory you could take are direct and bi-elliptic.  More wordage below, but the executive summary of these two paths looks like this:

• Direct path:  You eject from Kerbin in the solar-direction, and shoot out to the intercept in a fairly straight (i.e. only slightly curved) line.
• Bi-elliptic path:  The exact 180-degree opposite.  You eject from Kerbin in the solar- direction, such that your path takes you down to a fairly low solar Pe.  Then you wait until you get down to solar Pe and do a big burn there.  In other words, you eject from Kerbin going the "wrong direction" but then use the sun's gravity to make a U-turn.

In case you're wondering, "what's the longest I'd have to wait for one of these windows to pop up, some analysis in spoiler.

Spoiler

So, as you picture Kerbin orbiting the sun in its 360-degree circle, there will be some point on that circle that's the best place to eject for a direct path... and there will also be another point, roughly 180 degrees away, that will be the best place to eject for a bi-elliptic path.  Those two points will be the "best" places to eject.  The worst places to eject will be 90 degrees away from those points.  Since you'll design your ship to have a fair amount of extra dV, you're okay with being a reasonable distance away from the optimal points-- you just want to avoid the worst ones.

Let's say we're okay with launching when we're, oh, 20 degrees plus-or-minus from the "ideal".  What's the absolute longest you'd need to wait to launch?  Well, in the worst-case scenario, you've just gone a bit too far (i.e just over 20 degrees) past one opportunity, so you need to wait until you're 20 degrees before the next opportunity on the opposite side of Kerbin's orbit.  So, that means in the worst case you'd need to wait for 180 degrees of Kerbin's orbit, minus the 20 degrees you've already gone past one window, minus the 20 degrees before the next opportunity, i.e. 140 degrees of Kerbin's orbit.  So that's about 39% of an orbin, meaning you've got a reasonable window coming up no more than 160 or so days away even in the absolute worst case.

TL;DR:  you don't have to wait all that long for a window.  Significantly less than half a Kerbin year, even in the worst possible case.

Okay.  Now that we've established that you won't have to wait too long... let's figure out which of the two path types you want, and when to launch.

To describe these, I'll go back to the clock metaphor again.  Picture the solar system as a clock dial, with the sun at the center and your target intercept location (i.e. nearly out at the distance of Neidon) as being on the rim of the clock.  On that scale, Kerbin is orbiting really really close to the central hub of the "clock".  To keep visualization easy, let's rotate our frame of reference so that the intercept location is at 12 o'clock on the clock rim.  I'm assuming you're looking down at the solar system from over the sun's north pole, so all the planets are orbiting counterclockwise.

So what we're trying to figure out is, "at which o'clock would Kerbin need to be located in order to launch on a direct path?"  And same deal for a bi-elliptic path.  Got it?

Okay, let's look at the direct path first.

• Imagine the slowest possible scenario, where you're trying to optimize for "minimum dV" and don't care how long it takes to get to the destination.  In that case, you'd want to launch when Kerbin is at the 6 o'clock position.  That's because your transfer path is half of an ellipse, so you're going through 180 degrees of clock in order to get to the destination.
• On the other hand, imagine the fastest possible scenario, where you're trying to optimize for "get there as soon as possible" and you have infinite dV to play with.  You're shooting the ship like a rifle from a gun, and you'll be travelling in a nearly perfect straight line the whole way.  In that case, you'd launch when Kerbin is at the 3 o'clock position, because the target's at 12 o'clock and Kerbin-at-3-o'clock is where Kerbin is traveling in the desired direction, i.e. straight towards the target.

So, those are the two extremes.  You're going to be somewhere between those two extremes, because on the one hand, you do want to get there reasonably quickly (you have to catch up to the target, after all), but on the other hand, you don't have infinite dV.

So, as a broad guesstimate, you'd probably want to launch when Kerbin is somewhere between the 4 o'clock and 5 o'clock position-- I'd guess closer to 4 than 5.

Okay, that's the direct-path launch position.  What about the bi-elliptic launch position?  Well, it's around 180 degrees away, so that would be likely between the 10 o'clock and 11 o'clock positions.

Right.  So.  Rotate your map view so that the intercept location is at 12 o'clock, and look at where Kerbin is now, and picture Kerbin continuing in its orbit.  Will Kerbin arrive at roughly 4-5 o'clock sooner?  Or 10-11 o'clock?  That'll tell you which one of the two main path types will open up to you sooner, so you can pick that one.

 

Sorry... I know that's a whole lot of words.  (I tend to do that a lot.)  But the actual concept behind it is pretty straightforward, and hopefully it'll be clearer when you actually look at it in-game.

Anyway... how's this so far?  Does it make sense?  Is it useful?  Too much detail?  Too little?  Just wanted to touch base before going on to the tricky bits in step 3. 

[bigcalm]

Thanks all - sorry for the delay in replying, I've been away.

Snark - I think that should sort me out.  My plan, such as it is is as follows.

* As I'm around 130 days from the Neidon launch window, I'm going to wait a little while before sending the rescue, my thoughts being that Kerbin will be in roughly the same position and having similar orbital parameters to the Isaac Newton at that time.  That said, I'm going to stick it in orbit now and play with maneuver nodes.  The vessel I'm planning should have at a minimum 12k delta v, probably quite a bit more.  Note that the Isaac Newton isn't really on a Hohmann Transfer from Kerbin to Neidon - that would take around thirty years - so it's already pretty darned eliptical.  

* As a backup plan, I will try and rendezvous Nissee Orbit Spares with Isaac Newton around 7 years in.  I will wait until the correction burns have been done for both vessels first though.

* In the next Neidon window, I need to send a few extra bits, namely a replacement for the Isaac Newton (as the return journey won't be possible with its supplies of fertiliser).

This is all down to not checking properly during the design stage.  I only realised that it wouldn't make it when I came to design for Plock which is an even longer trip.  That said, I'm firmly of the opinion that Kerbal Space Program is at its absolute best when things go slightly wrong - not "smashing into the Moon at 500ms wrong", but where things don't quite work out as planned and there's a chance to fix it.

This is a long career with lots and lots of stuff going on.  I probably won't know if the above works for several real life weeks - but what you say sounds like it should work, as long as I pack in as much delta-v as I possibly can to the rescue ship I *should* be ok.

[JoE Smash]

I'm sure you are good at docking/rendevous in general (as you are doing advanced stuff and using life support mods in the first place), but as a noob who just managed to dock for the first time ever two days ago I wanted to pass on what I learned it the hard way.

In the event you have mechjeb installed ( I barely know how to use mechjeb either) under "RCS Balancer" make sure you have the RCS Overdrive slider down to 0 (well unless you need stronger RCS thrusters I guess). Mine was set to 100% by default and was making my first docking attempt seem totally impossible....as well as waste a crap-ton of monopropellant....

I'd tap the RCS thruster H key and have my target probe moving all over the place....

Once I turned that off, I couldn't understand why I was so scared of docking in the first place.

Good luck, it sounds like you have a lot of options and good advice here already! Without knowing what to do I'd just over engineer a rocket with the 6 tons of fertilizer, twice the delta v of fuel I thought I needed, and engines that were powerful enough to catch up and then cross my fingers and hope for the best....

[Snark]

7 hours ago, bigcalm said:

The vessel I'm planning should have at a minimum 12k delta v, probably quite a bit more.

Whoa.  Okay, you're loaded for bear.  Having that kind of dV budget means that you've got so much "wiggle room" that it's more like "wave your arms around freely and don't care if you smack someone" room. 

Anyway, sounds like you've got a handle on it, and no troubles in the engineering department.  I expect you'll be on track.  Good luck!

7 hours ago, bigcalm said:

That said, I'm firmly of the opinion that Kerbal Space Program is at its absolute best when things go slightly wrong - not "smashing into the Moon at 500ms wrong", but where things don't quite work out as planned and there's a chance to fix it.

Amen to that, brother!  I'm always secretly thrilled when something goes wrong (generally due to some stupid oversight on my part).  "Oh boy, now I get to figure out a way to pull this mission out of the hole!"  Those are some of the best moments in KSP. 

Spoiler

 

[Kryxal]

11 hours ago, bigcalm said:

This is all down to not checking properly during the design stage.  I only realised that it wouldn't make it when I came to design for Plock which is an even longer trip.  That said, I'm firmly of the opinion that Kerbal Space Program is at its absolute best when things go slightly wrong - not "smashing into the Moon at 500ms wrong", but where things don't quite work out as planned and there's a chance to fix it.

Actually, I had a mission where I did just that (okay, it was probably only 400 m/s) on takeoff and still made it back.