Deep Space Rendezvous

[justidutch]

I pulled off something the other day that I am extremely proud of, and thought I would share how it was done. It may turn out to be one of those things that I am ignorantly unaware as being common or easy, and if so I apologize for wasting your time. But I haven't been this proud of something since my first Mun landing over 18 months ago.

I'll spare you the details of why I got into this position, but if you're curious and have time to burn, the story is posted here.

The upshot of it all was that due to my own error in construction I found a ship's command module spontaneously disassembled from the rest of it, so it was dead in the water. Or space. Whatever. And, it was in deep space, its orbital path around the sun, periapsis from just outside Kerbin's SOI and apoapsis touching Duna's orbit. Basically in the middle of nowhere, no planets even remotely close by. I decided to try a rescue mission, so I built another ship that would attach a new command module via the Claw. But how the heck to do the rendezvous?

I had tried a similar deep space rendezvous about a year ago but gave up immediately when the closest I could get a rendezvous point was over 10,000 km. But I thought to give it a go this time. So, after getting the rescue ship into orbit and burning out of Kerbin's SOI into deep space, I set up a manoeuvre node and once again the closest I could get to the broken ship was ~12,000 km. Disappointing. But I thought maybe somewhere along the way I could refine it. So I time-warped about halfway, stopped, and rotated my ship to each of the 3 axis points. First I tried prograde-retrograde; small squirts of RCS thrusters forward and backward to see if I could get the encounter closer. And I could, by a little bit. The I tried normal-antinormal, that made a bigger difference. Then I did the same with the other axis, and that also made a slight difference. By the time that was done, the encounter was down to ~9,000 km. Better but still way too far away.

Timewarped halfway throughout the remainder of the course to the broken ship, stopped and made all the same adjustments. Then again and again and again. By the time I was three days away from the rendezvous I had whittled it down to 5 km!!! I honestly didn't think that was possible, but it turns out that it is.

So, the moral of my story is that no ship, no matter where it is (other than the surface of Jool I suppose) is without hope!

[Tex]

Excellent work! I've never tried anything so much as an interplanetary landing, but that is a wonderful accomplishment!

[martinborgen]

I found the rendevous system terribly borked - it always told me abot 60km, when I saw it passing it some 15km away...

[Grumman]

I found the rendevous system terribly borked - it always told me abot 60km, when I saw it passing it some 15km away...

I've noticed that too. The issue appears to be that the closest approach is always measured when your ship is at the point where the orbits come closest to crossing. But if you have a significant difference in velocity between the two ships, they might pass one another before or after that point, but be long gone by the time the other gets there.

[lazarus1024]

I don't think I've gotten within 5km before at such a distance, but I have gotten within something like 50km and tightened it up after that and gotten dock. As you found, you sometimes have to tweak several times on you way towards your meeting place. I doubt I've had the distance within 1,000km at anything over half an orbit away. Time warp to a quarter orbit, tweak a little. Time warp a little more, tweak a little more. Rinse and repeat.

You kind of, sort of have to do the same thing with asteroid rendezvous (unless you get one within an SOI of something).

Ignoring asteroids, I think I've done it 3 or 4 times over the years. Once to rescue a ship, once to try it and 1-2 times because I was building a space station orbiting between Duna and Kerbin. I only latched up 1 or 2 modules before I threw up my hands and said it was too much work.

[magnemoe]

I've noticed that too. The issue appears to be that the closest approach is always measured when your ship is at the point where the orbits come closest to crossing. But if you have a significant difference in velocity between the two ships, they might pass one another before or after that point, but be long gone by the time the other gets there.

Yes, I have done some interplanetary asteroid missions, tricks is to update nearest point as you get closer, and yes missing by 50 km is not an big issue, just burn towards target, time warp, at 10 km you tend to pass by the side so kill relative speed and repeat, do again at 2 km.

[Idrryl]

Next stop: solar station

[StrandedonEarth]

Congrats, deep space rendezvous is one of the tougher things to do. I remember how proud I was when I rescued Jeb from kerbolar orbit somewhere near Eve back in 0.23.

[luizopiloto]

I do this to get E class asteroids... to use as base for stations arround LKO...

[Skorpychan]

Next stop: solar station

I'm planning to do that in low solar orbit. But I'm planning to cheat, by just arranging all the modules evenly and thrusting it all out of Kerbin SoI together.

[Idrryl]

I'm planning to do that in low solar orbit. But I'm planning to cheat, by just arranging all the modules evenly and thrusting it all out of Kerbin SoI together.

Why not something like Kepler? To put it in the same orbit like Kerbin, only slightly slower/higher.

[Caelib]

Good job, this is definitely a challenge! In my experience, you're hitting an unpredictable moving target with floating point calculations ... there were many times where I tried to do a rendezvous, but when I would time-warp, the target would be at a completely different distance than what was displayed prior to time-warp.

Moral of the story: time-warp + deep space rendezvous == baaaad

[KrazyKrl]

Even with a 12,000km closest approach, there is an easier way to get to your rendezvous. When you start approaching near a closest approach; start pushing your retrograde indicator to your anti-target indicator, what this does; is it moves your orbit vector to nearly identical to your targets. Your objective in a rendezvous situation is to make your orbits identical, and as such; all you need is a reasonably small percentage of your altitude as closest approach distance. Just plop down a maneuver node at the closest approach to get an idea of how long the burn will be to get a reasonably close orbit; and plan your available burn time accordingly.

A side effect of pushing your retrograde indicator onto your target indicator; you are burning all of your deltaV to get that approach closer; and is more efficient in most cases than random orbital burns.

Also, don't get hasty with the final approach; small amounts of timewarp are your friend, and even a small 50m/sec target speed is more than plenty with timewarp.

Edited May 26, 2015 by KrazyKrl

[AbacusWizard]

Well done! Few things in KSP are quite as satisfying as a successful long-shot rescue mission.

[Raptor9]

Even with a 12,000km closest approach, there is an easier way to get to your rendezvous. When you start approaching near a closest approach; start pushing your retrograde indicator to your anti-target indicator, what this does; is it moves your orbit vector to nearly identical to your targets. Your objective in a rendezvous situation is to make your orbits identical, and as such; all you need is a reasonably small percentage of your altitude as closest approach distance. Just plop down a maneuver node at the closest approach to get an idea of how long the burn will be to get a reasonably close orbit; and plan your available burn time accordingly.

A side effect of pushing your retrograde indicator onto your target indicator; you are burning all of your deltaV to get that approach closer; and is more efficient in most cases than random orbital burns.

Also, don't get hasty with the final approach; small amounts of timewarp are your friend, and even a small 50m/sec target speed is more than plenty with timewarp.

^^^This...angle of approach applies to interplanetary transfers, you want that shallow orbit intersection so you're not trying to change direction AND slow down.

@justidutch, congrats on the successful rescue mission

[mrclucks]

I find interplanetary roundevouz pretty easy. The hardest part is the amount of time needed and matching speeds