Achieving a perfect orbit

[A1catraz]

Every time I launch a vessel into orbit, I can never get he Apoapsis and Perapsis perfectly matched, it's always a couple ten or hundred off. Is perfect orbit possible, if so, how is it achieved?

Edited November 8, 2013 by A1catraz

[jduchock]

the definition of peri and apo are "closed" and "farthest" from a given body, respectively...

If the peri and apo were the exact same number, they would cease to be peri- and apo- apsises...

the apoapsis will always be a higher value than the periapsis...

[Pursuedtank]

Realistically an orbit with the same apoapsis and periapsis is very difficult to do, and as soon as you time warp or exit to space center it will change ever so slightly. Slightly off in a magnitude of a few tens of meters is perfectly fine. You don't need a perfect orbit usually anyway.

[TelluriumCrystal]

Probably not, considering KSP uses floating point numbers and that the Apoapsis and Perapsis markers start flipping out exponentially the closer they get.

If you really want a perfect orbit, you only real solution is to edit the persistence file. Unfortunately I cannot offer any advice on that subject.

Hope this helped.

[Smidge204]

In practice, +/- a few percent is pretty good. More than good enough for anything you might want a circular orbit for.

As was said already, if you get the numbers really close (Like +/- 500 meters out of 100km!) the markers glitch right the f*ck out due to floating point errors. This will not help you.

Squad should bite the bullet and implement their own big number library already stupid Unity engine...

=Smidge=

[Seret]

Is perfect orbit possible, if so, how is it achieved?

Any orbit where you manage to not hit the big round bit in the middle is good enough IMO. You don't get any extra space points for perfection, close enough is close enough. It's up to how OCD you are about how perfect you want to try and get it. I generally consider any Kerbin orbit within about 2km variation to be circular enough for me.

[capi3101]

Hell, I'm pretty sure any orbit with a 2k variation is good enough for NASA...yeah, they don't get it exactly right either. That includes their geosynchronous satellites too, BTW; they occasionally require adjusting to keep them in the correct position tolerance.

You're okay if the two are within a klick of one another for pretty much anything you want to do. The only place I've seen really close precision count has been in forum challenges (and not all that often).

[TMS]

Every time I launch a vessel into orbit, I can never get he Apoapsis and Perapsis perfectly matched, it's always a couple ten or hundred off. Is perfect orbit possible, if so, how is it achieved?

Simple way is to 'chase' the apoapsis along your velocity vector when you have an ETA of 1 second, pushing your periapsis towards an equal value. It'll get you as close as possible. RCS will help you fine-tune.

But yeah, it's not desirable (or, by definition, achievable) due to the errors mentioned earlier, and it making maneuver placement difficult.

[Kasuha]

When the orbit is too circular, weird things start happening. The game couldn't decide where AP/PA points are and you may have very hard time placing a maneuver on it. It's better to stick to very slightly elliptical orbits.

Apart of that, the less parts and the finer control your ship has the more precise you can make the orbit. I would imagine using a probe core, a large RCS tank and two RCS ports for positioning might allow you to make orbit circular within 1 m.

[ultimatevirus]

My best was about <1 km accuracy. My only real advice would be use more then 1 or 2 burns, maybe like 5 or 6 using low thrust engines so that you can tune the level of thrust easily.

[ThirdHorseman]

As everyone else said, it's pretty impossible. Which is tough, because I tried to setup two matching refuel stations around Kerbin, each at 150k and orbiting opposite each other. Of course since you cannot get an exactly circular orbit they tended to get closer and closer together over time so I abandoned that idea...

[EtherDragon]

It is possible to achieve a circular orbit, just very very difficult, and there is no benefit to doing so. In fact, you use a fair bit of propellant trying to tweak your orbit to get there. (An orbit who's Periapsis and Apoapsis are at the same distance is just a circle...)

[DisarmingBaton5]

People on this thread: The weird oscillations in your orbit about the apoapsis and periapsis are caused by the motion of your command pod. It is not random.

[Gnipu]

To fine tune orbit: Go to an -apsis, align with it, hit Caps Lock and fire RCS forward og backwards and watch the opposite -apsis change.

[RadHazard]

If you want to sync two or more satellite or station orbits so they stay roughly the same distance apart, it's easier to tweak the duration of each orbit than it is to make them perfect circles. You'll need a mod that displays more orbital info for that though.

[Mr. Entropy]

If you can get the apoapsis and periapsis within 2 km of each other, that's good enough for government work.

[Zuni]

Quick answer: Mechjeb.

[TelluriumCrystal]

Quick answer: Mechjeb.

Not even Mechjeb can achieve a perfect orbit. I speak from experience.

[Smidge204]

People on this thread: The weird oscillations in your orbit about the apoapsis and periapsis are caused by the motion of your command pod. It is not random.

It's absolutely true that applying torque tot he ship can cause the Ap/Pe points to jostle around a bit, but with orbits of very, very low eccentricity they will glitch out even with all reaction wheels and SAS disabled. Two different causes contributing to the same effect.

=Smidge=

[basdxz]

The most precise orbit in my experience can be achieved using ion engines. They produce very little thrust making them slow, but dead accurate.

[Kerbart]

If you can get the apoapsis and periapsis within 2 km of each other, that's good enough for government work.

I think most of us have higher standards than that.

[Starwhip]

Use RCS Thrusters. They are fairly precise. I can get within 100-500 meters of equilibrium.

[Spyritdragon]

Use the blue thingumats on maneuver nodes. If youre right between apo and peri, one direction will increase apoapsis and decrease periapsis while the other does the opposite. But as said before, within 1km is usually good enough

[A1catraz]

Didn't really expect this many replies, anyway, thanks for the help, I just wanted to know if it was possible. It obviously is not. What could be the closest to a "perfect orbit"?

[Ralathon]

Didn't really expect this many replies, anyway, thanks for the help, I just wanted to know if it was possible. It obviously is not. What could be the closest to a "perfect orbit"?

I think I got it as perfect as possible in the game while building my Astraeus constellation:

The eccentricity of all those orbits is down to 0.0000XX according to flight engineer and the same is true for the relative inclinations. They all have a orbital period of exactly 1.5 hours. Each orbit contains 3 sats at 120 degree angles.

At this point the apoapsis, periapsis and ascending/descending node are jumping around like crazy. Every single timewarp or reload shifts your orbital parameters. Therefore I gave up trying to perfect it, it would be pointless as a single reload would mess it all up again.

To do this to this degree of accuracy you need some form of flight computer. I used the remotetech2 flight computer for this, but I'm guessing Mechjeb or kOS works just fine too. You need the flight computer so you can exactly point your ship in any of the 6 directions. It is hard enough to get these numbers down without a burn on the inclination adjusting the period.

In general: Burn prograde/retrograde to increase/decrease period. try to cancel vertical speed at the halfway point between apo and peri to circularize and fix your inclination as you always would. When the numbers get low enough you start doing really small puffs of thrust, tap shift and tap x immediately afterwards. If you have access to a flight computer that can que engine commands it helps a lot to make some sort of macro that burn the engine for 0.01 sec at 1% thrust. Using these methods you can usually make the orbit perfect in about 2/3th of an orbit, just in time to deploy the next satellite if you're doing a constellation like mine.