What is the most efficient way to make large orbital plane changes?

[NJC2]

EDIT: I posted a followup question here.

Original post:

---

Long Story (skip to TL;DR if you want the what and not the why): I'm still fairly new at this. So far I've just done low Kerbin orbit stuff. I wanted to see if I could figure out how to get into orbit by myself without reading any tutorials. Having only done the in-game tutorial, I set a goal for myself to get a satellite in orbit. I somehow managed to do it on my first try, but it ended up in a wild elliptical retrograde orbit and offset 25° (it's on a heading of 245°). Also once I put the satellite in orbit I was unable to get the rocket stage out of orbit. At the time I didn't realize I needed both battery power and a command pod/module to control the rocket stage after detaching the satellite. It took me three more attempts before I successfully put a satellite in orbit and de-orbited the rocket stage(s). After that first attempt I decided to read some guides on the wiki. As it turns out, the ship I built happened to stage exactly with the basic tutorial on getting to orbit. The first stage burned out right at 10km where I pitched over to 45° until I was at 70km, then burned prograde until I had an orbit.

That first attempt was sloppy, and I just got lucky. After reading the tutorial, I did the rest of my launches making my gravity turn to the east. As I said previously, it took me another three attempts to get a satellite in orbit and not leave the rocket stage up there too. By then I had accumulated three junk rocket stages and one non functional satellite (I forgot to deploy the solar panels before its battery ran out). I decided to build a craft to see if I could clean up my orbital junk rather than delete it from the Tracking Station because that felt like cheating. Initially I was going to see if I could do it consistently, then I would just delete it from the Tracking Station and pretend I sent up a "deorbiter" to do it.

It turned out to be quite a challenge and each one was different, and since it wasn't that much junk I decided to go ahead and manually do them all. The first three craft were single use designs. For my first design, I decided to go after the non functional satellite. I put LT-1 lander legs on the tip of the craft thinking I would use the gear to snag the junk, and perform a deorbit burn. However there was a bug in them at the time. I had to quit before completing the mission, so I saved the game and quit. When I came back and continued, the landing legs broke off after loading it up. I was so close to intercepting the satellite, so I decided to continue and see if I could just bump into it and nudge it out of orbit. It worked, so my second design I just swapped out the lander legs with an upside down small to large adapter thinking the larger surface would make it easier to bump things. I think I just got lucky with that first one, because the second one was harder. I went for a junk rocket stage, and I had a tough time with it. The junk kept rolling off when I fired the engines, and I had to re-capture it three times before I slowed it down enough for it to deorbit.

For the 3rd try I went back to using lander legs. This time I designed a kind of basket, and used the LT-2 lander legs as a lid. I went after another junk rocket, and it worked out great. After deorbiting the junk rocket, there was plenty of fuel, RCS and battery left to go after another piece. I decided to go ahead and deorbit one of the satellites since they aren't really doing anything useful. It was when going after the satellite that I realized there was a flaw in the design. The smaller satellite could slip out between the beams. I had to re-capture it a few times, but I was able to get it out of orbit.

After getting two pieces of junk with that last design, I decided to build a permanent orbiting craft that I could just re-fuel and keep re-using. I decided to make it modular by putting a large clamp-o-tron on the end, so I could make different attachments and swap them out. I basically just made a larger version of the last design, but put a docking port for refueling and solar panels to recharge the battery. It had the same flaw as the previous design that would let small objects slip out since the struts don't collide with objects, they slip right through. I still used it to de-orbit all but two pieces of junk, the first satellite I launched and its rocket that are in the bad orbit. I made a new module, sent it up and attached it, then refueled, and now I need to match orbit with those last two pieces of junk.

TL;DR: To get the last two pieces of space junk I need to make a large orbital plane change of ~155°. I tried playing with maneuver nodes to see what would work the best, but I couldn't find a way to do it with them since I figured I would need to make several maneuvers instead of one big one. At first I thought I should be able to do it by burning 90° between the ascending and descending nodes, but I couldn't get that to work with a maneuver node. So instead I made smallish adjustments (about 1-2min burns) at each intercept node. I've gone from a 90°E orbit to almost a polar orbit of about 170°, and now I'm out of fuel. I'm going to send up a refueling ship and continue. At this rate I should get it matched to the orbital plane of the junk, but it will need another refueling once I get there before I can capture and deorbit the junk.

There must be an easier, more efficient way than what I'm doing. As much fuel as it's going to take to get there doing it the way I am, it would have been more efficient to just launch another ship directly into that orbit.

I had a few ideas of what to do with it once I finished cleaning the junk. One was to just leave it where it is until I need it to clean up more junk. Also, since it's modular and with lots of fuel, I could always use it as a rescue vehicle in case I strand some Kerbals somewhere. All I would have to do is send up a crew module or lander module to attach to it depending on where they are stranded. Another idea I had was use it as the start of a space station. For either of the last two options I first need to return it to an equatorial orbit. But first I have to finish the job it was intended for.

Edited November 23, 2013 by NJC2
Marked as Answered

[DevoidLight]

What I do is set the Mun as my target, because it's on the same plane as Kerbin's equator, and burn at the ascending/descending nodes that show up.

[Geschosskopf]

Significant inclination changes while in low orbit are always going to cost an insane amount of delta-V. There's no way around it. It's usually pointless to try unless you're using an ion- or VASIMR-powered thing with 10s of thousands of delta-V, or the infinite fuel cheat.

However, even with all the delta-V in the world, making the burns while still close the planet just don't work very well. The burns are going to be long even for chemical rockets and close to the planet, you're moving so fast you zip by the node before you can do much good. So, the easiest way to make a big plane change accurately and all at once is to do it far from the planet. This dramatically reduces the delta-V required for the plane change itself, so that even a xenon-burner can handle it in short time, which allows for fine control. HOWEVER, you have to burn a lot to move your Ap out this far, and then after the plane change you have to bring your Ap back in. So, you don't really save any fuel doing it this way, you just gain better control over what you're doing.

[DMagic]

HOWEVER, you have to burn a lot to move your Ap out this far, and then after the plane change you have to bring your Ap back in. So, you don't really save any fuel doing it this way, you just gain better control over what you're doing.

You might be able to save some fuel if you can use the atmosphere to bring your Ap back down. There's is probably some critical altitude that you have to achieve to balance the delta-v required for the inclination change with the delta-v required for raising your Ap. But I would imagine that planets with an atmosphere could tilt this trade off in favor of raising your Ap first, rather than trying to do it all in low orbit.

[capi3101]

Geschosskopf has given you good advice - what he's describing is the first phase of what's called a "bi-eliptic transfer", where you kick your apoapsis high enough that it takes relatively little delta-V to make changes at the periapsis. Massive orbital plane changes is one of the uses of such a transfer; another is to get in close to Kerbol or Moho. You know that low delta-V value that appears on the delta-V map for a Moho transfer? It assumes the bi-elliptic.

The trade-off: you use less delta-V overall, but it takes significantly longer to perform (the Hohmann to Moho takes 5,000-7,000 m/s and about 40 days; the bi-elliptic takes no more than 3,000 but it will take close to 250 days just to reach the apoapsis). So take to your pick as to what you want to do.

[arq]

As the others are suggesting, burn so that your apo is out around Minmus. Make sure that your apoapsis is at/near the ascending/descending node of your target. At the AN/DN out near Minmus, change your orbit to whatever inclination you want, then drop your periapsis down into the atmo a little (maybe 35km? i dont remember) so that it will bring your apo back down.

[KSK]

Rather than trying to correct your inclination once your in orbit, you could try launching directly into the right (or approximately right) orbit to start with. It'll take more fuel since you can't take full advantage of Kerbin's rotation to help you get to orbit but that should pale into insignificance compared to the amount of fuel you save by not having to make that big an inclination change on-orbit.

Caution - this next bit is more speculative.

To do this, I think you'll need to:

1. Have your deorbiter on the launch pad and ready to go.

2. Switch to your space junk and locate the ascending and descending nodes on its orbit.

3. Timewarp until the launch site is a few degrees behind either of the nodes. Having the deorbiter sitting on the pad makes this easier as the launch site becomes more visible. Basically you want to time things so that your deorbiter hits that node as it reaches orbit.

4. Switch back to your orbiter. Go to map screen and select the junk as a target to give you a target cue on your nav-ball.

5. Launch and follow that target cue to orbit!

6. Good luck!

Incidentally - respect for spending all this effort to deorbit your space junk!

[Geschosskopf]

I'd forgotten that you could aerobrake at Kerbin. I've been doing mapping probes all week at airless bodies with the frequent need for major inclination changes.

As I recall, a good altitude to aerobrake at Kerbin if you 1) are starting about as far out as Minmus and 2) don't want to land, just reduce your Ap, is about 37-40km. Much lower and you land, much higher and it'll take numerous passes. As always with aerobraking, F5 beforehand so if it doesn't work for you, you can F9.

[Sirrobert]

Land and set a new craft up, with the correct inclination

[SRV Ron]

If you can, set up a slingshot with a moon;

[DrMonte]

Significant inclination changes while in low orbit are always going to cost an insane amount of delta-V. There's no way around it. It's usually pointless to try unless you're using an ion- or VASIMR-powered thing with 10s of thousands of delta-V, or the infinite fuel cheat.

However, even with all the delta-V in the world, making the burns while still close the planet just don't work very well. The burns are going to be long even for chemical rockets and close to the planet, you're moving so fast you zip by the node before you can do much good. So, the easiest way to make a big plane change accurately and all at once is to do it far from the planet. This dramatically reduces the delta-V required for the plane change itself, so that even a xenon-burner can handle it in short time, which allows for fine control. HOWEVER, you have to burn a lot to move your Ap out this far, and then after the plane change you have to bring your Ap back in. So, you don't really save any fuel doing it this way, you just gain better control over what you're doing.

No, you can save a significant amount of fuel this way. See Oberth effect for more details here.

Involving Mun and aerobraking you might get where you want. But I'm not sure if you have to go so far.

[Kasuha]

The most efficient way to perform such a large orbital plane change (when not using gravitational assist of other bodies) is bielliptic transfer. Raise your apoapsis, perform the plane change at the apoapsis, then lower the apoapsis again.

[Kasuha]

Caution - this next bit is more speculative.

To do this, I think you'll need to:

1. Have your deorbiter on the launch pad and ready to go.

2. Switch to your space junk and locate the ascending and descending nodes on its orbit.

3. Timewarp until the launch site is a few degrees behind either of the nodes. Having the deorbiter sitting on the pad makes this easier as the launch site becomes more visible. Basically you want to time things so that your deorbiter hits that node as it reaches orbit.

4. Switch back to your orbiter. Go to map screen and select the junk as a target to give you a target cue on your nav-ball.

5. Launch and follow that target cue to orbit!

6. Good luck!

1. Have your deorbiter on the launch pad and ready to go

2. Switch to Map view and select the junk piece as your target

3. Time warp until the orbit goes above KSC

4. Launch in the direction of the target orbit. Target markers can provide you rough direction but do not burn directly at them, rather just the same direction from vertical (and make sure you don't get on retrograde orbit relative to your target!)

5. rendezvous and do whatever you need to do

[Frederf]

There's a handy graph around that shows how much you want to puff up your orbit for various inclination changes. The bigger the change the more you want to increase your orbit. I can't find the graph though.

[Geschosskopf]

No, you can save a significant amount of fuel this way. See Oberth effect for more details here.

This is a very common misconception amongst KSP players. Oberth does you no good at all if you stay in a closed orbit the whole time. Oberth only makes a difference to flybys, not closed orbits around the same planet, and even then only for chemical rockets with high fuel flow rates. All Oberth does is increase a rocket's fuel efficiency the faster the rocket is moving. It does not create free energy and it doesn't help low-flow engines like ion drives much if at all.

In the case of a closed orbit, where you do 3 burns to change your inclination, Oberth's effects cancel out. While Oberth makes the 3rd burn more fuel-efficient than the 1st burn thanks to your higher speed at Pe, you only have that higher speed because you did the 1st burn. Thus, the apparent savings on the 3rd burn are illusory because you paid for them up front. Instead of both burns using the same amount of fuel, the 1st burn uses more than the 3rd because it is RELATIVELY less efficient, but they still add up to the same amount.

[Nao]

There's a handy graph around that shows how much you want to puff up your orbit for various inclination changes. The bigger the change the more you want to increase your orbit. I can't find the graph though.

This awesome program has the graph you are looking for.

Bi elliptic transfer starts being better from singe burn plane change from 38,94deg up to 60 where leaving SOI and returning with proper inclination would be better (at lest that's how i understand Bi-parabolic transfer in KSP ).

[numerobis]

Set up a Mun fly-by; if you set your periapsis somewhere other than the equator, it'll provide you an inclination change for free. Except that you flew out to the Mun.

Really the cheapest way is to fix your inclination on launch.

[NJC2]

Thanks for the ideas everyone. I didn't think about extending my Ap and making the maneuver farther out. I knew what I was doing couldn't have been very efficient since the burn to change the plane also increased the orbit and I was wasting fuel to prevent my orbit from growing larger at the same time. I did actually think about slingshotting out around the Mun and coming back, but that seemed extreme to me at the time.

[Nao]

Slingshotting around Moon to change inclination was actually done in real life too, (i don't remember exactly) but i think some satellite was placed in reverse or polar orbit through moon flyby because they couldn't launch it from available sites into desired orbit. I think it was because of safety reasons as most launch centers have some kind of unpopulated area on the early ascent path (water usually) to the standard orbits, and launching in other direction meant flying over populated areas so direct launch was out of the question.

[NJC2]

I haven't had the time to get back to KSP since my last post, but today I'm hoping to finish my junk cleaning mission. I just refuled the deorbiter and the refueler has about 2/3 of its Rockomax x16 tank left, so I'm going to take advantage of it to push the deorbiter.

Now I'm ready to finish the orbital inclination change. I pulled up this thread to re-read before starting the maneuver. Is it really necessary to go as far out as the Mun or Minmus for an efficient inclination change? I was thinking I could put the ship into a synchronous orbit, turn it in the direction it needs to go, then burn to establish the orbital intercept. I know this would require burning up to a synchronous orbital altitude, another burn to slow down, then another burn to get the intercept orbit, but would it use more fuel than a large burn to go out as far as the Mun or Minmus and back?

[tstehler1]

Slingshotting around Moon to change inclination was actually done in real life too, (i don't remember exactly) but i think some satellite was placed in reverse or polar orbit through moon flyby...

Let's also not forget about Apollo 13. They pulled a bit of a slingshot maneuver as well.

[NJC2]

I decided to give my synchronous orbit idea a try, and it worked out quite well. Rather than actually put the craft into a synchronous orbit, I just pushed my Ap up to the synchronous orbit altitude (~2,870km since I didn't need to be exact), then made my inclination change burn. I was able to make the last ~70° of inclination change to match orbit with the debris, drop down to intercept and de-orbit the first piece. I then re-establish orbit, intercepted the second debris and de-orbited it. I had about 1/4 of the Rockomax x32 tank left, and was curious if I could get it back to equatorial orbit with that. I pushed my Ap back up to to synchronous altitude (except I overshot and ended up just above 3,000km). I had plenty of time to play around with the maneuver node, and I managed to use the node and plot the 155° inclination change, keeping roughly the same Ap and Pe all in one burn, and it was less than a 1000m/s burn! I didn't quite have enough fuel to pull it off, but I was able to use the RCS to at least maintain orbit. When it was all over with I was only 5° off from an equatorial orbit.

My Pe was at 80km and Ap was still around 3000km. At Pe, I burned more RCS to lower my Ap down to around 2500km, then decided to try aero braking. I've never done aero braking before and I didn't know the best altitude to use, so I went for 50km. I wish I'd thought about it before and planned the maneuver node to set it up from the start. The aero braking was working, but I was only losing 25km per orbit and that was going to take too long to circularize my orbit that way, so put my Pe up to 75km then burned up the rest of my RCS and was able to lower my Ap down to 1200km. I'll send up fuel tomorrow, fix the orbit and then park it for whenever I need it later.

I'm pretty sure that wasn't the most efficient way to do it because I don't think it even takes a 1000m/s burn to reach the Mun. I quick saved before I did it because I didn't believe what the maneuver node was telling me, and didn't think it was going to work. Once I saw that it actually worked, I figured what the hell and went with it.

[John FX]

I decided to give my synchronous orbit idea a try, and it worked out quite well. Rather than actually put the craft into a synchronous orbit, I just pushed my Ap up to the synchronous orbit altitude (~2,870km since I didn't need to be exact), then made my inclination change burn. I was able to make the last ~70° of inclination change to match orbit with the debris, drop down to intercept and de-orbit the first piece. I then re-establish orbit, intercepted the second debris and de-orbited it. I had about 1/4 of the Rockomax x32 tank left, and was curious if I could get it back to equatorial orbit with that. I pushed my Ap back up to to synchronous altitude (except I overshot and ended up just above 3,000km). I had plenty of time to play around with the maneuver node, and I managed to use the node and plot the 155° inclination change, keeping roughly the same Ap and Pe all in one burn, and it was less than a 1000m/s burn! I didn't quite have enough fuel to pull it off, but I was able to use the RCS to at least maintain orbit. When it was all over with I was only 5° off from an equatorial orbit.

My Pe was at 80km and Ap was still around 3000km. At Pe, I burned more RCS to lower my Ap down to around 2500km, then decided to try aero braking. I've never done aero braking before and I didn't know the best altitude to use, so I went for 50km. I wish I'd thought about it before and planned the maneuver node to set it up from the start. The aero braking was working, but I was only losing 25km per orbit and that was going to take too long to circularize my orbit that way, so put my Pe up to 75km then burned up the rest of my RCS and was able to lower my Ap down to 1200km. I'll send up fuel tomorrow, fix the orbit and then park it for whenever I need it later.

I'm pretty sure that wasn't the most efficient way to do it because I don't think it even takes a 1000m/s burn to reach the Mun. I quick saved before I did it because I didn't believe what the maneuver node was telling me, and didn't think it was going to work. Once I saw that it actually worked, I figured what the hell and went with it.

To use the least fuel you should have reduced your Pe to about 45km (maybe a little less for the first one) and kept aerobraking and adjusting your Pe and you can get the Ap you want with good accuracy. An aerobrake at 60km+ will only drop your Ap by a few km then when your Ap is right (at 150km say) do a small burn at Ap to raise your Pe over 69078m

[maltesh]

This is a very common misconception amongst KSP players. Oberth does you no good at all if you stay in a closed orbit the whole time. Oberth only makes a difference to flybys, not closed orbits around the same planet, and even then only for chemical rockets with high fuel flow rates. All Oberth does is increase a rocket's fuel efficiency the faster the rocket is moving. It does not create free energy and it doesn't help low-flow engines like ion drives much if at all.

I'm afraid you're incorrect here. The Oberth effect occurs regardless of how efficient your engine is, or whether or not your orbit is closed. It doesn't increase the rocket's fuel efficiency at all. A rocket will still have the same amount of delta-V, regardless of how fast its going. By expending that delta-V when the spacecraft is moving quickly, the same amount of delta-V is better at increasing orbital energy, and that orbital energy is what determines how fast the rocket winds up going at any altitude it winds up reaching, as well as the semimajor axis of the resulting orbit.

[Meithan]

Significant inclination changes while in low orbit are always going to cost an insane amount of delta-V. There's no way around it. It's usually pointless to try unless you're using an ion- or VASIMR-powered thing with 10s of thousands of delta-V, or the infinite fuel cheat. [...] HOWEVER, you have to burn a lot to move your Ap out this far, and then after the plane change you have to bring your Ap back in. So, you don't really save any fuel doing it this way, you just gain better control over what you're doing.

Nope, that's not correct. You can save delta-v by doing the inclination change at a higher altitude, even including the delta-v spent in the two burns required to first raise and then lower back the apoapsis, and without aerobraking. Here's an example with numbers.

Suppose we're in a 80 km circular orbit around Kerbin, and we want to change our inclination by 90°. We have two options.

Option A: perform the inclination change without changing orbit

Computing the required delta-v is straightforward here.

Delta-v required: 3223 m/s

Option B: raise our apoapsis, do the plane change, then lower the apoapsis back to 80 km

Here we have to calculate the delta-v in three steps.

1) First, raise the apoapsis to, say, 1000 km altitude.

Delta-v for ap increase: 421 m/s

2) Perform the inclination change at new apoapsis (1000 km)

Delta-v for inclination change: 1623 m/s (quite a lot less!)

3) Burn at periapsis to circularize back to 80 km

Delta-v for circularization: 421 m/s (the same as the first, of course)

The end result is the same.

Total delta-v required: 2465 m/s

That's a sizeable saving!

And the more you raise the apoapsis, the larger the saving (although you start getting diminishing returns as you increase the apoapsis more and more).

With aerobraking you can cut down on that third delta-v too.

Edited November 24, 2013 by Meithan