Off-world Refueling: A short explanation

[Northstar1989]

I've seen a number of questions by new players asking for advice from "experienced players" about the benefits of refueling depots and stops on the way to places like Jool and Eelo vs. straight transfers from Kerbin.

There is a wide spectrum to the term "experienced player", and different players have different skill-sets and specialties. Speaking from the perspective of an experienced player who does a LOT of refueling, gravity-assists, and aerobraking; all to save fuel and Delta-V, I thought I'd try and clear this topic up a little for new players. Sorry if it drags on a little- I first wrote this at roughly 2:30-2:50 AM, though I've done my best to clean it up since...

TECHNICALLY, it costs more Delta-V to do a refueling stop, and then continue on your way to your destination, than just to make a straight transfer, all other things being equal- but there are several EXTREMELY important exceptions to this that can make it a lot more worthwhile:

(1) You can save a lot of fuel if the body you make your refueling stop around has an atmosphere. This may not seem intuitive, but aero-"braking" can actually be used to ACCELERATE your ship towards your destination. Let me explain this:

-Take the example of Duna, for instance. If you set up a fuel depot in low orbit around Duna, and send a spacecraft from low-altitude orbit around Kerbin to Duna via a low-energy transfer; Duna will be traveling a lot FASTER relative to the Sun than your spacecraft by the time it gets that far away from the Sun... Now, when you enter Duna's SOI, you will have a large velocity difference, and appear to be traveling very "fast" relative to the planet- but from the point of reference of the Sun, it is actually DUNA that is traveling very fast relative to YOU. If you dip into the atmosphere around Duna, it will exert acceleration on your spacecraft to bring it closer to Duna's speed. In essence, you are performing an aero-boost, so to speak, relative to your final target Jool.

- Now, if you got your fuel into Duna orbit in the same way (low-energy transfer and aerocapture) you will have saved quite a bit of fuel, as it is approximately 900 less Delta-V to make an ideal transfer to Duna than to make an ideal transfer to Jool (this ISN'T counting fuel-savings from a Munar gravity assist- which will save you close to the same amount of Delta-V on either trip if you perform it right). Now, it does cost approximately 1350 Delta-V to make a transfer from circular low Duna orbit to Jool- but you can make up quite a bit of that extra 450 Delta-V with a gravity-assist from Ike. Additionally, your approach speed at Jool will be lower when transferring from Duna instead of Kerbin (so it will take a lot loss fuel/aerobraking to capture) and you can make a MUCH more precise burn from Duna than from Kerbin (the burn-time will be shorter), which will more than make up the rest of that Delta-V in terms of reduced steering losses, reduced approach velocity, and less need for course corrections (especially if you are using NERVA engines...) Of course, this is where (2) comes in...

- For BONUS fuel-savings, don't perform an aerocapture to Duna orbit- perform one to Ike. It's a bit tricky to pull off, but if you set you aerobrake such that your exit trajectory comes out with an apoapsis appropriately-oriented to Ike, you can perform an easy capture there, and save a lot more Delta-V than from capturing to Duna orbit- as it takes less Delta-V to make it from Ike orbit to Jool if you make appropriate use of the Oberth effect... (see below)

(2) Craft design and the Oberth Effect:

- The first, and most important thing you need to know about efficient spacecraft design is, the fuel-efficiency of your craft is *STRONGLY* connected to its fuel fraction. That is, it's not efficient to push around a bunch of empty fuel tanks or excessive numbers of decouplers. One of the main reasons off-world refueling is profitable is because it allows you to maximize the fuel-fraction of your spacecraft without a bunch of staging (which is both expensive and heavy- a 2.5 meter inline decoupler weighs and costs a LOT more than a Clamp-O-Tron docking port. And radial decouplers tend to be even worse...) Creating a partway stop for refueling along your mission route enables you to consolidate the number of vessels moving along the route from a larger into a smaller number as fuel is consumed- even if it's just a single fuel tanker and your interplanetary mission ship. Docking and fuel-consolidation has the same effect as staging: but is usually much lighter and cheaper. (you can dock two 100-ton vessels with nothing but a Clamp-O-Tron Jr: try using THAT to connect two 100-ton fuel sections in your mission vehicle, and see what happens...) Setting up a fuel-depot is of course heavy and expensive, but that's more of a one-time investment that you make for convenience, and isn't strictly necessary- your fuel tanker(s) *can* wait on their own at the refueling stop, and hook up with your mission vehicle directly.

- Second, getting an accurate transfer burn to Jool from low Kerbin orbit is hard. More likely than not, you will need to perform a long, slow burn around Kerbin; followed by a shorter adjustment burn somewhere in interstellar space (MechJeb can help you find the optimal location for these adjustment burns, to minimize Delta-V costs. You could theoretically determine this yourself, but the mathematical equations required are nothing to sneeze at, and it will take a LONG time to locate the optimal adjustment points if you utilize trial-and-error course-plotting instead...) The fact is, the shorter a burn you have to make to get somewhere from Kerbin, the more accurate it is likely to be- simply by virtue of gravity having less time to curve you off-course during your transfer-burn. Additionally, there is something important you need to know about called the "Oberth Effect"...

The Oberth Effect is a mathematical phenomenon that is EXTREMELY important to orbital mechanics. Without going into too much detail about it, you should know the basic result of it: the faster you are going when you burn your engines, the more energy you get out of it. What this basically means, in game-terms, is that it takes a LOT less Delta-V to get somewhere from a departure-point where you are already moving fast than in does to get there when you start out moving slowly. For this reason, you cannot simply find the difference between your current velocity and the velocity you need to be going at when you arrive at your destination to calculate Delta-V costs. The larger the Delta-V of a transfer burn from planetary orbit, the relatively less Delta-V it costs you compared to the final total kinetic and potential energy you will have at your final destination- as you will be moving quite quickly towards the end of your transfer burn. If you make a transfer burn from low above Kerbin, it costs you less Delta-V than if you first raise your orbit to the edge of Kerbin's SOI, and then make the transfer from there. It is actually for this reason that an *ideal* straight transfer (with no steering-losses: though such burns are, in practice, *impossible*) to Jool from Kerbin *theoretically* costs you less than a burn to Duna and then from there to Jool, assuming you didn't make use of Duna's atmosphere to provide you with a portion of your velocity. However, the Oberth Effect leads to one other important point you need to know about in order to understand why refueling (at Duna or anywhere else) can actually make sense:

+ Due to the way the Oberth effect works, the higher your acceleration is relative to the local gravity when you make a transfer burn, the less Delta-V it is going to cost you. This is actually for a rather simple reason, really- if your acceleration is much lower, your burn will be much longer, and you will make a greater portion of it in a more distant orbit from the celestial body you are departing. This means your average velocity during the burn will be lower (the further you are from a body, the lower your orbital velocity). Put another way, you have already lost a portion of you velocity from the beginning of your burn to gravity by the time you are nearing the end of your burn. I don't know how to explain this better- the greater your TWR, the less Delta-V you will spend reaching your destination, even before you account for the greater steering-losses to compensate for gravity-turning due to longer-burn time that a lower TWR also inevitably entails...

+ For obvious reasons (Duna's gravity is less than 1/3 Kerbin's), the same exact craft will have a much greater TWR around Duna than around Kerbin. What this means is, you will come a LOT closer to the idealized transfer of 1350 Delta-V from low Duna orbit to Jool's SOI than you will to the idealized transfer of 1915 Delta-V from low Kerbin orbit to Jool (you will lose less of your velocity to gravity over a burn of the same length- and thus your average velocity will be higher). Put simply, when you combine this with reduced steering-losses it's not 1915 Delta-v vs. 1060 + 1350 Delta-V (idealized transfers), it's more often 2100 Delta-V vs. 1150 + 1420 Delta-V. Just from this fact alone you will save some Delta-V with the same craft design, even before you account for the additional gravity-assist that Ike offers you the opportunity to make (from Duna orbit), or the opportunity to aerobrake directly into an Ike-intecept, and refuel around Ike rather than around Duna, which will make your Delta-V savings even greater...

- I mentioned craft design earlier, and this is really another important factor to consider. In order to make a reasonably-accurate burn to another planetary system, you generally want to keep your burn-times to no more than 5-6 minutes (shorter if your departure orbit around Kerbin is REALLY low- I usually depart from at least 100 km or higher in order to make my transfer burns more accurate, and give my ships a little more breathing room to make rendezvous with fuel tankers or space stations for fuel, and sometimes crew, transfers before departure). Consider the following:

+ A craft with a TWR of 1 around Kerbin can accelerate at a little under 10 m/s^2. To make a transfer to Duna at that TWR would take a little over 2 minutes (assuming an expenditure of 1200 m/s Delta-V with steering-losses). For the same craft to make a transfer to Jool it would have to burn for at least 4 minutes (assuming an expenditure of almost 2400 m/s to Jool). This is ignoring the effect of staging and increasing TWR with fuel-losses of course, but in reality the TWR of NERVA upper-stages is also likely to be quite a bit lower...

+ Let's say your craft is of such mass as to have a TWR of 0.25 around Kerbin with a single LV-N (NERVA) engine. To raise that to 1.5 would take MORE than 6 LV-N's- as each additional engine adds a not-insignificant mass to the vessel. Additionally, each additional LV-N also reduces your available Delta-V budget for the same reason.

+ The factor of craft structural stability and design-flexibility also comes into play. To make a reasonably-accurate burn to Duna, you won't need as high a TWR as to make an accurate burn to Jool. This means your craft will be placed under fewer Kerbin-G's of acceleration for the transfer burns if you make the refueling stop at Duna (the aerobraking, on the other hand, might place your craft under quite a few G's- but this step isn't strictly-required), or better yet, Ike (if you can, *ALWAYS* make the refueling stop at Ike, never make the stop at Duna- you will save a lot more Delta-V this way if you can still pull it off with an aerobrake, and especially, if you can't- say if your craft is too flimsy to survive aerobraking...) This gives you more options for craft-design: a craft that doesn't need to pull as many G's, and can get by with less thrust, has a lot more options for its design than one than needs to cram in 6 or 8 LN-N engines... You can even utilize some of the more thrust-anemic, but higher-ISP options than NERVA's this way, especially if you are using mods: for instance KSP-Interstellar's ATTILA engines, or NearFuture pack's HydrogenGas plasmodynamic thrusters (which are basically upsized ion engines that utilize a much less dense reactive propellant in the form of hydrogen gas, rather than Xenon, but benefit from improved ISP *and* TWR as a result- compared to stock ion engines...)

+ For those of you wondering how you could get a craft into orbit around Kerbin than couldn't survive at least 1 Kerbin-G of acceleration, there are two answers to this question: the first is that even with stock gameplay, you can construct large multi-part vessels through docking in-orbit that are often much more flimsy as a whole than their individual modules. The second answer is that with modded gameplay, certain mods (such as Orbital Construction and Extraplanetary Launchpads) actually allow you to construct spacecraft in-orbit out of parts (resources) that you ultimately hauled from the ground on Kerbin or produced from locally-available resources off-planet.

- Finally, it should be obvious to you that there is an upper-limit to the size of your spaceships. This is determined both by the limits of the game itself (i.e. part-count and strain on the physics engine), and by your ability to construct a structurally-sound vessel that isn't so complicated as to cause even you, its designer, to forget how to use it properly. Personally, I use mods like NovaPunch2 and StretchyTanks to scale up the maximum size of my craft (by providing larger-diameter parts, I can build larger vessels with reasonable part-counts), and sometimes I really push the limits of part-count sanity nonetheless; but I still make use of refueling stops to allow me to bring larger vessels to more distant destinations... (in fact, I often utilize refueling around the Mun or even Minmus just to reach Duna with many of my craft designs- usually because of one design constraint or another, such as their ability to act as a high-performance plane on Duna, or a frankly insane 240-ton payload size meant to help me establish a large colony on Duna that I simply couldn't build a big enough rocket to move to Duna without refueling...)

(3) In-Situ Resource Utilization:

- I'll be brief on this one. It (hopefully) goes without needing too much explanation that it is much lighter, and thus costs a LOT less fuel from Kerbin, to send a craft to another celestial body (such as the Mun or Duna) to produce and refine resources there (such as Kethane, or KSP-Interstellar's ability to perform electrolysis of Muna regoltih for Aluminum and Oxidizer), than it does to simply send all the fuel you need from Kerbin. In order to make use of these resources, you're going to either need to make refueling stops at the production location, or send fuel tankers from the production location to somewhere less out-of-the-way for your interplanetary vessel. Often, you'll make use of a combination of the two- for instance mining regolith on the Mun or Kethane on Duna, and then launching the resources to a fuel depot in orbit- which is the actual vessel your interplanetary ship will dock with and refuel.

- With enough mods installed, you can produce resources virtually anywhere. The five most important resource-utilization mods I know of are KSP-Interstellar, Kethane Pack, Karbonite, Extraplanetary Launchpads, and (indirectly) NearFuture mod. KSP-I offers a wide variety of resources in appropriate locations, ranging from Munar regolith to LFO mix (presumed to represent combustible ratios of cryogenically-stored Hydrogen and Oxygen gas) extracted from electrolyzing water from Vall's (presumed) ice deposits. Kethane offers, well, Kethane- wihch can be refined directly into LFO, Monopropellent, and even Xenon. Karbonite does pretty much the same thing, but with a different (slightly better) system. Extraplanetary Launchpads offers a three-step resource system in which you mine Ore, then refine it into metal, then again into RocketParts, and finally use a craft-construction part on an existing base or vessel to construct additional ships like at the VAB or SPH, except off-world. Finally, there is NearFuture mod. It doesn't directly give you any ISRU capabilities- but its *HUGE* solar panels (ranging as large as the newest, 250 EC/s, "solar blankets") are the only *practical* way to make large-scale use of KSP-Interstellar's beamed microwave-power system to create solar electricity-farms in orbit.

- It also goes without saying that In-Situ Resource Utilization is the *ONLY* way you will ever set up sustainable long-term operations on other planets. It simply takes too many fuel tanker launches to send all the fuel you need to do all the things you want to do otherwise- and it's going to get even worse when they introduce in-game economy, and you have to pay for all these fuel tanker launches as well. I especially recommend Extraplanetary Launchpads and Orbital Construction mods for this purpose (please note, Orbital Construction does not provide a way to manufacture RocketParts on its own- you'll need to also install Extraplanetary Launchpads if you want this functionality- and in fact Extraplanetary Launchpads offers its own orbital-construction mechanic that is slowly making Orbital Construction mod obsolete, if not for the much larger RocketParts "warehouse" modules it offers...)

Regards,

Northstar

P.S. As always, reputation points are appreciated if you found this thread helpful. You can award it using the little star symbol in the lower-left corner...

EDIT: Updated explanation to include mention of Karbonite mod as a possible ISRU solution

Edited October 4, 2015 by Northstar1989

[Northstar1989]

It's also been pointed out to me an alternative strategy for making maximal use of Duna's atmosphere when trying to reach a more distant location such as Jool, or Sentar (an even more distant gas giant with rings- a Saturn analog from Krag's Planet Factory) to save fuel. This works best in combination with In-Situ-Resource Utilization, as it costs quite a bit of fuel for the refueling base on Duna...

(1) Instead of aerocapturing to a low orbit, or to Ike orbit, aerocapture to a highly elliptical orbit with an apoapsis close to the edge of Duna's SOI. DO NOT unthinkingly make your aerocapture into a prograde orbit relative to Duna's rotation, or you might end up waiting a longer time to make your eventual ejection burn. Make your atmospheric pass on whichever side of Duna is closest to the sun (the orbital direction will thus vary depending on your approach trajectory- this aerocapture will place you into a retrograde orbit if approaching from behind Duna relative to its orbit, but a prograde orbit if approaching from in front...)

(2) Raise the periapsis of your orbit above the atmosphere. You won't want an additional atmospheric pass slowing you down. If you have enough fuel, this is also a good time to make a phase shift to your orbit by making a low-intensity burn radially outward from Duna at apoapsis, or radially inward at periapsis. Your phase shift should change the major axis of your obit so it points more in the direction of Duna's movement relative to the sun.

(3) To refuel the spacecraft, send a tanker from Duna or Ike orbit. It will cost quite a bit of fuel for the tanker, but will save fuel on the ejection burn of the main vessel vs. aerocapturing to a lower circular orbit around Duna or Ike.

(4) Eject the newly refueled interplanetary vessel from Duna's orbit, making the majority of the burn at periapsis with Duna to make maximal use of the Oberth Effect. For bonus fuel savings, and awesomeness, grab a gravity assist from Ike. The major axis of your elliptical orbit will be pointed mostly radially outward away from the sun, and perhaps also a bit retrograde relative to Duna's orbit around Kerbol, so you will need to either perform a phase shift first to shift the direction of your major axis (burn radially outward near apoapsis, or radially inward near periapsis- though be careful, it's easy to accidentally introduce a component in other directions to your burn during the phase shift, creating accidental undesirable orbit changes), or correct for the degree to which your major axis is off by introducing an extra prograde or radial inward/outward component to your ejection burn.

The main benefit of this method is, as you can see, is that it saves you a LOT of Delta-V on your ejection burn (up to about 500 m/s worth- if you remember, the base cost of the two part transfer is only 450 m/s more if you make the 2nd burn FROM LOW DUNA ORBIT- so you make up the cost difference this way even before you account for some of the other sources of fuel savings such as shorter and more accurate ejection burns...) HOWEVER, it can take a very long time compared to other methods, is rather complex, and can easily be messed up by unplanned encounters with Ike.

Thus, I don't particularly recommend this method, but only thought I'd put it out there for any of you gutsy enough to give it a try- or who had heard of it before, and were curious about some of the details on how to actually do it (which, even now, I've painted in rather broad strokes...)

Regards,

Northstar

Edited June 22, 2014 by Northstar1989

[Northstar1989]

I made some corrections to the numbers in this thread, and updated the explanations a bit. I hope it's still useful to you guys!

Regards,

Northstar

[J3ansley]

This is fantastic. I'd give you rep but I'm too new to.

I may have missed it, but on fuel critical missions do you prefer Mun over Minmus or vice versa?

[Northstar1989]

This is fantastic. I'd give you rep but I'm too new to.

I may have missed it, but on fuel critical missions do you prefer Mun over Minmus or vice versa?

I don't think I explicitly addressed that- so you didn't miss anything.

On critically fuel-limited missions (such as when moving extremely heavy payloads, where it's just not possible to cram in enough fuel tanks for a robust Delta-V budget), the best strategy is to perform what I like to call an "Oberth Dive" from Minmus.

Basically, it works something like this:

(1) Set up a refueling depot in low orbit around Minmus.

(2) Refuel your interplanetary vessel there

(3) Perform an ejection burn with your interplanetary vessel from Minmus orbit such as to return towards Kerbin with a low periapsis BUT NOT ENTER THE ATMOSPHERE. How low of a periapsis is fuel-optimal depends on how far away your target is from Kerbin Delta-V wise (and how fast of a transfer trajectory you opt for- if you opt for speed over fuel-efficiency).

(4) At or near periapsis with Kerbin, perform your transfer burn towards your intended destination (obviously, this requires some planning in advance such that your burn/periapsis is at the correct Angle to Prograde)

The large fuel-savings from both the Oberth Effect and the high velocity you will already have when beginning your ejection burn make this method extremely effective. Note that you will still need a substantial TWR for an accurate ejection burn (lower TWR will require a large adjustment burn later- though this may be worth the reduced engine mass, and corresponding increase in Delta-V) despite the small Delta-V of the ejection burn, as your Angular Velocity will be VERY high at periapsis, leaving you a very short window to make an accurate burn.

For even more fuel-critical missions (I've utilized this method once before when I was sending a SSTDABK-capable spaceplane to Duna on a very suboptimal transfer window) you can set up a full fuel tanker in the same orbit around Minmus, but at a slightly later phase- such that there will be enough time to switch to it and perform an identical ejection burn from Minmus when in reaches the same Angle to Prograde as your interplanetary vessel made its ejection burn...

You then have the fuel tanker match the transfer burn near Kerbin, and rendezvous with the interplanetary vessel as it is on its outbound trajectory to provide the necessary fuel for the adjustment and capture burns at the destination (any fuel needed for the return voyage can be more easily/cheaply provided be fuel tankers in orbit of the destination- which can be sent either ahead of or behind the interplanetary vessel, with much lower TWR).

Note that this will require substantial skill/knowledge in both orbital maneuvering (MechJeb will help, but even then you need to know how to make use of some of the more advanced maneuvers it can plot) and rendezvous, as well as as a high TWR on the fuel tanker to help it make an accurate burn to follow/rendezvous with your outbound interplanetary vessel.

The actual docking is no more difficult than one in a normal orbit though- simply match velocities at closest approach (by burning along the retrograde vector of relative velocity in "Target" mode), get closer, match velocities again, and repeat ad-infinitum as usual until you're close enough to dock the two vessel with a simple burst of thrust (from the frame of reference of the fuel tanker/ interplanetary vessel, the two vessels will be stationary relative to each other when their velocity is matched and distance is small, even if hurtling towards the edge of the solar system at thousands of meters/second).

Regards,

Northstar

Edited July 8, 2014 by Northstar1989

[Torrack]

Northstar,

If you don't yet work for NASA, you should. I'm fairly certain that was the best "down to Kerbin" explanation of Orbital mechanics I've read in quite some time. Well written and thought out; thank you for posting!

[Illumia-Star]

Why can't I rep more than once?!

This is going to be so helpful! Thanks a bunch!

[Northstar1989]

Why can't I rep more than once?!

This is going to be so helpful! Thanks a bunch!

Glad you like it. With the addition of Budgets in 0.24, off-world refueling just became a LOT more useful- as it turns out the devs opted for the realistic/accurate approach, and made fuel tanks much more expensive than the fuel they contain. Decouplers, structural parts, and even struts are also fairly expensive- so there's an exponential increase in the cost of progressively larger rockets.

Being able to launch a larger number of smaller to mid-sized rockets (which can also thus rely more heavily on cheap SRB's to get off Kerbin), and use some of them as fuel tankers, is looking to be a lot cheaper than single heavier launches...

Not to mention the benefits of reusable rockets and spaceplanes for filling up fuel depots around Kerbin- and reusables don't scale well to ultra-heavy payloads (like interplanetary transfer missions) due to the need for stability when landing back on the ground...

Regards,

Northstar

Edited July 19, 2014 by Northstar1989

[Frannington]

Same here. This has been very useful to me - thank you.

[Decent Weasel]

Okay, two questions-

1) how does a simple slingshot around the back of a body compare to using aerocapture to add velocity?

2) how would TWR factor into maneuvers made in freefall? Isn't efficiency due to Oberth related to the gravity gradient, which would favor higher gravities?

[mhoram]

1) how does a simple slingshot around the back of a body compare to using aerocapture to add velocity?

EDIT: See explanation below for a more fitting answer.

Aerocapture decreases the velocity of a ship at the location where the ship is while flying through atmosphere.

Slingshots only work, if SOI-changes are involved.

While orbiting Kerbin you can use the Mun to change the orbit around Kerbin.

There is a detailled description of Gravity assists and a graphical description available.

2) how would TWR factor into maneuvers made in freefall? Isn't efficiency due to Oberth related to the gravity gradient, which would favor higher gravities?

TWR tells you how long you need to perform a burn. So it affects the time needed for a burn.

And yes Oberth around Jool can be much more effective than Oberth around Dres.

Edited September 26, 2015 by mhoram
typo

[Northstar1989]

Aerocapture decreases the velocity of a ship at the location where the ship is while flying through atmosphere.

Read the explanation carefully. Aerocapture increases your velocity relative to the *SUN* in the situations described (at the terminal end of a transfer orbit), NOT relative to the planet.

The planet is moving faster than you relative to the Sun when you are in its SOI- and your velocity relative to the planet is actually its velocity relative to you (from the frame of reference of the Sun, you're moving slow and the planet is moving fast). So anything that decreases your velocity *relative to the planet* in that situation increases your velocity relative to the Sun.

Slingshots only work, if SOI-changes are involved.

While orbiting Kerbin you can use the Mun to change the orbit around Kerbin.

I'm not sure what you're trying to say here. You can make a slingshot relative to the Sun around any planet orbiting it, given the right velocity and angle of approach (conversely, you can also perform a gravity *brake* instead of an assist, if you pass the planet on the opposite side from an assist...) Gravity slingshots are NOT an artifact of the Patched Conics method of predicting trajectories used by KSP- they're a thing in real life. In fact, NASA used a whole bunch of them for the Voyager and Galileo missions:

http://www2.jpl.nasa.gov/basics/grav/primer.php

Regards,

Northstar

[Northstar1989]

1) how does a simple slingshot around the back of a body compare to using aerocapture to add velocity?

It depends a LOT on the specifics. Theoretically, a slingshot can be a lot more effective- it can add up to twice the orbital velocity of a planet around the Sun or a moon around a planet to your spacecraft, *if approached from the correct trajectory and angle*. However, achieving the angles and velocities necessary relative to the Sun for that kind of an assist are so difficult and/or expensive in terms of Delta-V that you're MUCH better off with an aerocapture.

In practical terms, an aerocapture is much easier and should be more than enough. For instance, if you aerocapture into an elliptical orbit around Duna with a periapsis *just above* the edge of Duna's atmosphere (you'll need to perform a *slight* prograde burn at apoapsis to achieve this periapsis), and an apoapsis at the very edge of the SOI; then you could theoretically gain up to the escape velocity of the planet (which for Duna is 1372 m/s) towards your next leg of the journey further out, if your target was far enough out that your escape burn ensured you lost almost no velocity on the way out from Duna... (this would mean a burn for a transfer to a different *star system* however- in practical terms, your gains are more likely going to be limited to around maybe 700 or 800 m/s)

However, as I stated before, aerobraking at Duna to get to Jool would actually cost *more fuel* if you looked strictly at Delta-V costs (you would have to travel to a destination significantly further out than Jool for a Duna aerobrake to save Delta-V directly: think a trip to the Sentar system in Krag's Planet Factor...) The main benefit is that you can perform a much more ACCURATE pair of transfers this way than if you just went directly to Jool- which means you can use lighter engines (which increases your total Delta-V budget) and subject your rocket to fewer G's of acceleration. And the equation changes ENTIRELY if you're using ISRU- in which case the total rocket mass you need to launch from Kerbin orbit becomes MUCH smaller (than a direct transfer, or refueling around Duna using tankers) if you can build a reusable fuel infrastructure around Duna...

2) how would TWR factor into maneuvers made in freefall? Isn't efficiency due to Oberth related to the gravity gradient, which would favor higher gravities?

Oberth efficiency is related to gravity gradient in not one but two ways:

The first is, of course, the potential amount of fuel you can save using the Oberth Effect. That is positively related to the strength of the gravity field, and thus the velocity of your starting orbit. The faster your orbital velocity, the more energy you gain for each each second your engines are firing...

The second is how much of that potential you can actually reasonably harness with a given rocket design. This in negatively related to the strength of the gravity gradient. The stronger gravity is, and the less sharply it falls off as you leave the planet (a larger planet with a lower density and the same surface gravity is more challenging in this regard- hence one of the many reasons Real Solar System is so much harder than stock despite Earth/bigger Kerbin having EXACTLY the same surface gravity as stock Kerbin...), the more kinetic energy you'll lose from the start to the finish of the burn. This is due to two factors: stronger gravity fields are less forgiving of long burn times, and your burn time will need to be longer in order to escape the gravity field in the first place.

The first of these there is nothing you can do to change, other than launching from a lower orbit around a more massive body. The second, however, you can somewhat counteract by giving your rocket more powerful (and thus heavier) engines, and performing one or more "periapsis kicks" so more of your burn is performed at greater orbital velocities...

Burn time basically is negatively related to your TWR around the body you are orbiting, and the stronger the gravity, the lower your TWR with that planet, and the longer your burn time for the same rocket design. So a two-part journey that makes a pit-stop around Duna has the advantage of shorter ejection burn times BOTH due to the lower Delta-V requirements for each burn, AND the higher TWR your rocket will have with the same engines when orbiting Duna due to the weaker gravity...

Regards,

Northstar

Edited October 13, 2014 by Northstar1989

[mhoram]

Aerocapture increases your velocity relative to the *SUN* in the situations described (at the terminal end of a transfer orbit), NOT relative to the planet.

Interesting concept. Never thought about Sun-relative velocities while aerobraking - I always had only the velocity within the target SOI in mind.

This is also the reason why my other statement does not fit as an answer to Decent Weasels question.

[Decent Weasel]

Fascinating... thanks guys!

[Kulebron]

Great idea with aerocapture. I could not wrap my head around this instantly. Question is, if you're on a highly elliptical trajectory around Duna, what's the departure window to Jool? Because if the ellipse is pointing the wrong way, it's expensive to change the direction. Or do you just depart from Duna, having already gained ~500 m/s?

Also, my concern is that launch window for Duna are every ~300 days, and Duna=>Jool are ~1000 days.

[Matrix Aran]

Great idea with aerocapture. I could not wrap my head around this instantly. Question is, if you're on a highly elliptical trajectory around Duna, what's the departure window to Jool? Because if the ellipse is pointing the wrong way, it's expensive to change the direction. Or do you just depart from Duna, having already gained ~500 m/s?

Also, my concern is that launch window for Duna are every ~300 days, and Duna=>Jool are ~1000 days.

This is a problem with all gravity assists, and one of the reasons we haven't been darting around the solar system as much as we'd like. As an example, the alignment of planets that made the Voyager grand tour possible is an event that only occurs once every 175 years. However, this simply encourages extra creativity on the part of the folks designing a trajectory.

As a general rule of thumb though, any trajectory that saves you deltaV is likely to cost you time, with a few exceptions. Say Jool isn't in the right phase for a direct transfer, but it just so happens Duna is available. If your desire is to launch right now, you can take advantage of Duna for a fuel savings and get off the launchpad before Kerbin pulls your funding for wasting the snack budget on another munar hop.

[sal_vager]

Great thread Northstar1989, stickied for November

[Northstar1989]

Great thread Northstar1989, stickied for November

Awesome!

What an honor! I might have to go back to the OP, and clean up the grammar and syntax a little more. Maybe clarify a couple of the more ambiguous parts of the explanations...

Regards,

Northstar

[Northstar1989]

I just wanted to thank all you guys for all the love this thread has received! (tons of Rep points given, etc.)

It's been my pleasure to help you guys learn more about the value and logistics of off-Kerbin refueling!

Regards,

Northstar

[Northstar1989]

Updated this to include mention of Karbonite as a possible ISRU solution.

If any players notice anything else I missed or that has changed since I originally wrote this back in February 2014 and expanded it in May/July 2014, feel free to let me know!

Regards,

Northstar

[xtoro]

Just wanted to throw out another option that I don't think was mentioned, which I used myself: Using an unmanned refueling ship for a large "mothership".

Here's my refueling ship. It can take 2 karbonite tanks with it, or 1, or none. This gives flexibility to land on moons to mine and convert depending on the gravity of the body. Lower gravity moons can use both karbonite tanks. It will land, mine and convert into fuel. Again, if the gravity is low enough, not only can it refuel itself, but it can also take the full tank(s) of karbonite back up to the mothership and convert into fuel once docked.

It's not very large and doesn't hold a whole lot of fuel, but even with several trips, it can refuel my orbiting mothership and the landers on it. Is it efficient? Yes and no. It does the job of keeping the mothership refueled and ready to head to the next planet/moon, but it can be wasteful in regards to spent fuel to land and take-off every time. But, that's only if you consider the amount of karbonite being extracted from the planet. If you run this mission over and over and over, yes, you will eventually sap all the moons dry of fuel. But for a few missions, this works quite well.

Here it is being towed behind the mothership (between the large boosters).

Here is the whole mothership will all its landers docked. The mining lander is between the large boosters (left side of pic) which are only used to get to Eve.

So this is just another alternative to setting up large fuel depots for long missions. Just take a small mining/refueling rig with you and refuel on the way! This is what I'm doing with my mission/story in my signature.

[Northstar1989]

Just wanted to throw out another option that I don't think was mentioned, which I used myself: Using an unmanned refueling ship for a large "mothership".

Actually, that's *exactly* the kind of thing I meant by mid-mission refueling allowing use of ISRU. Cool demonstration of the concept though- rock on!

Regards,

Northstar

- - - Updated - - -

Is it efficient? Yes and no. It does the job of keeping the mothership refueled and ready to head to the next planet/moon, but it can be wasteful in regards to spent fuel to land and take-off every time. But, that's only if you consider the amount of karbonite being extracted from the planet. If you run this mission over and over and over, yes, you will eventually sap all the moons dry of fuel. But for a few missions, this works quite well.

Actually if you're taking your refueling rig along with you like that, smaller rigs are more efficient despite having to make more trips- because they require less fuel to get out to the planet/moon in the first place. In the long run, larger rigs are more efficient because they have relatively less mass in probe cores, landing legs, batteries, etc. and more in fuel-tanks: but for a mobile operation like that smaller rigs are much more efficient...

Also, Karbonite was undepletable last I checked. Did they change that recently? Because honestly, having depletable resources makes zero sense- the scale you are going to be extracting resources on is going to be *miniscule* compared to the mineral reserves or atmosphere or whatnot of even a local resource deposit on a planet/moon... It would be like going to an oil field on Earth and having it run out after just a couple days for resources to deplete on the tiny scale of a space program...

EDIT: Nope, they didn't change that. It still says right in the release thread's OP: "Karbonite's resources are inexhaustible. Kethane's resource deposits can be depleted."

So, either you're mixing up Karbonite and Kethane, or there's something wrong with your version of Karbonite, if your resources are being depleted...

So this is just another alternative to setting up large fuel depots for long missions. Just take a small mining/refueling rig with you and refuel on the way! This is what I'm doing with my mission/story in my signature.

It pays off to set up permanent outposts though. This doesn't even have to mean an actual depot in orbit or a refinery on the surface or in orbit. Simply sending an extractor-lander like the one you just showed to each planet and leaving it for the next mission (and the one after that, and the one after that...) will save you fuel if you carry out enough missions... But I'm sure you're aware of that!

Regards,

Northstar

Edited March 17, 2015 by Northstar1989

[eonmoon]

i can see a refinery being plopped down, a base built on several worlds to gather resorces, soon kerbals will be able to colonize the solar system ( plan for me later, tho once 1.0 comes out everything will change since karbonite will switch to stock resorces for the most part

[MatBailie]

i can see a refinery being plopped down, a base built on several worlds to gather resorces, soon kerbals will be able to colonize the solar system ( plan for me later, tho once 1.0 comes out everything will change since karbonite will switch to stock resorces for the most part

Does anyone have an opinion one what their "idealised" mining station would look like?

My first attempt had one ISRU and one Drill-O-Matic, but the ISRU was only being utilised to a fraction of a percent (partly low yield spot, partly no engineer, mostly that the ISRU has a massive capacity compared to a drill).

So my second attempt has 8 Drill-O-Matic to get some where near 30% utilisation on my ISRU on Minmus (highest deposit I could find was 2.7%), but that is totally covered in 16 Gigantor Solar Arrays (8 pointing horizontally over the Drills and two pairs of 4 extending vertically around the core's ISRU, holding tank, etc).

Neither has much in the way of batteries; I'd need so many to keep it running in the dark that it seemed more efficient to have two stations on opposite sides of the moon!

Then I found that even at 30% utilisation my ISRU can fill fuel tanks faster than the missions I launch uses the fuel. Now I'm thinking that I should ignore the ISRU utilisation and just accept it will always be more than I'd ever need? Or maybe the ISRU should be in orbit with a small fleet of mining vessels? (Which I avoided so as not to worry about running out of fuel before the mining vessel docked with the ISRU station.)

I'm also new to surface bases, so I steered clear of having a permanent base plus a refuelling ship. Instead my entire base returns to orbit to dock with an orbital station or orbiting vessel, and then descends again to mine some more. Maybe that's a "mistake" on my part?

What are other people's opinions? (I'm sure there are no "right" answers, I'm just intrigued by the different considerations people are taking in to account and how they go about addressing those.)

Mat.