Impossible contract? Satellite positioning close to Mun orbit but opposite direction

[THX1138]

Edit: I forgot to emphasise the fact that the orbit direction is against that of the Mun so every orbit has two encounters!

I just want to confirm before I give up that this mission is indeed impossible. The screenshot shows my attempt to put a satellite in the orbit specified (green). It was fairly close until a Mun encounter threw it off. I'm pretty sure that this orbit will never be stable because Mun encounters will always send the satellite off somewhere.

(click to enlarge)

Edited December 23, 2014 by THX1138

[Red Iron Crown]

It doesn't have to be stable, you just have to deliver it successfully to the specified orbit and you'll complete the contract. That it gets gravity assisted out of specification shortly after will not prevent contract completion.

[SRV Ron]

Contracts will specify for how long that orbit needs to be stable. Once completed, you can either let the satellite get kicked out or orbit on its own, or send it crashing into Mun to clear space.

[cicatrix]

... but reaching that orbit with Mun's getting in the way can be troublesome

but doable.

[Sirrobert]

... but reaching that orbit with Mun's getting in the way can be troublesome

but doable.

Not that troublesome. Just send it while the mun is on the oposite side as it would be if you actually tried to get to the Mun

[cicatrix]

Not that troublesome. Just send it while the mun is on the oposite side as it would be if you actually tried to get to the Mun

Who thinks of that BEFORE launch?

[Sirrobert]

Good point

[THX1138]

If you get one half of the orbit correct and then rotate around to the other side to get that half correct you'll meet the moon somewhere in between and be thrown off. I think the only chance of success is if you deliberately have bad ascending/descending nodes. After getting the apoapsis and peripasis correct, assuming you can get the ascending/descending nodes to zero in a single move then you can do it, notwithstanding a chance encounter near the ascending/descending nodes. Either way, this is surely unintended and should be fixed so these contracts don't appear.

[Stargate525]

If you get one half of the orbit correct and then rotate around to the other side to get that half correct you'll meet the moon somewhere in between and be thrown off. I think the only chance of success is if you deliberately have bad ascending/descending nodes. After getting the apoapsis and peripasis correct, assuming you can get the ascending/descending nodes to zero in a single move then you can do it, notwithstanding a chance encounter near the ascending/descending nodes. Either way, this is surely unintended and should be fixed so these contracts don't appear.

Assuming you've got a single intersection with the desired orbit, you can correct the orbit entirely with one burn at that point. It's more fiddly, but it can be done.

[Sirrobert]

If you get one half of the orbit correct and then rotate around to the other side to get that half correct you'll meet the moon somewhere in between and be thrown off. I think the only chance of success is if you deliberately have bad ascending/descending nodes. After getting the apoapsis and peripasis correct, assuming you can get the ascending/descending nodes to zero in a single move then you can do it, notwithstanding a chance encounter near the ascending/descending nodes. Either way, this is surely unintended and should be fixed so these contracts don't appear.

Not if you leave at the right time.

The Mun won't catch up to you in a single orbit if you arive in the orbit just behind it, giving you more than enough time to get the orbit right

[Red Iron Crown]

Hard to tell if it is so from the pic, but it might be more difficult if the desired orbit is retrograde.

[THX1138]

Hard to tell if it is so from the pic, but it might be more difficult if the desired orbit is retrograde.

It IS! That's the whole point! You cannot avoid constantly whizzing passing the damn moon! I can't believe I forgot to make that clear in my original post!

Edited December 23, 2014 by THX1138

[Red Iron Crown]

Well, that changes things a lot. Not sure if the contract's requirements will be met by merely being in an orbit with the correct parameters regardless of later encounters.

[hyperbolicj]

Well if the orbit is in fact in the muns' Sphere of Influence. Then remember that the orbit has to be stable for 10 seconds before the contract is finished. So the real objective is to send the satellite in such that you have the most possible time before the muns' SOI shifts the orbit due to an encounter.

Since the orbit is closer to kerbin your satellite will be traveling faster. Therefore the one thing we want to avoid is coming up from behind on the mun. So the ideal case scenario is to send the probe just ahead of the muns' SOI. That gives you the most time to send the rocket into the appropriate orbit before it comes in contact with muns SOI by catching up to it from behind.

[Anquietas314]

Well, that changes things a lot. Not sure if the contract's requirements will be met by merely being in an orbit with the correct parameters regardless of later encounters.

Nah that's completely fine. I've had a few contracts to deploy satellites in a (prograde) orbit that quite quickly has an encounter with the Mun. You obviously won't have as long to get into the right orbit though in a retrograde one.

[MrCynical]

Is there actually any check in the satellite orbit contracts that the orbits are possible? I've just had a really awful one to put a satellite in an orbit round Duna, which was basically a slightly inclined retrograde version of Ike's orbit. As far as I can tell there was no way to have a satellite in that orbit without having an encounter with Ike within half a lap. I eventually got it to just barely trigger as complete, but even that was clipping through the sphere of influence of Ike part way round (I'm not entirely sure whether it was checking the periapse value before or after that encounter, but they were close enough it was probably within the deviation allowed).

[sal_vager]

So it's a short lived satellite and also an Ike impactor? It needed a seismic accelerometer then

[Anquietas314]

So it's a short lived satellite and also an Ike impactor? It needed a seismic accelerometer then

Psst. You don't want to put the seismic accelerometer on the thing you're using as an impactor

[RocketPilot573]

The probe only needs to be there for 10 seconds....

[Anquietas314]

The probe only needs to be there for 10 seconds....

Yep. There's a whole discussion in the suggestion forum over whether it should be longer or how those should be changed, but currently 10 seconds is all you need

[MrCynical]

The probe only needs to be there for 10 seconds....

Yes, but you still need to match the orbit the specified, which is kind of difficult when there's always going to be an encounter somewhere between apoapse and periapse to throw it off. I'm not sure how it handles future encounters when determining whether a satellite is in the correct orbit. Even if it ignores them, there's still the problem that the displayed orbit will be distorted by the encounter - no easy way to tell what the hypothetical one would be if the moon wasn't there.

[Anquietas314]

Yes, but you still need to match the orbit the specified, which is kind of difficult when there's always going to be an encounter somewhere between apoapse and periapse to throw it off. I'm not sure how it handles future encounters when determining whether a satellite is in the correct orbit. Even if it ignores them, there's still the problem that the displayed orbit will be distorted by the encounter - no easy way to tell what the hypothetical one would be if the moon wasn't there.

If the shape of the orbit matches up pretty close, you know the periapsis must be close to correct, assuming you're burning at apoapsis.

[RocketPilot573]

Yes, but you still need to match the orbit the specified, which is kind of difficult when there's always going to be an encounter somewhere between apoapse and periapse to throw it off. I'm not sure how it handles future encounters when determining whether a satellite is in the correct orbit. Even if it ignores them, there's still the problem that the displayed orbit will be distorted by the encounter - no easy way to tell what the hypothetical one would be if the moon wasn't there.

You don't need to be at apo or pe to change the orbit. Try burning radial.

[MrCynical]

If the shape of the orbit matches up pretty close, you know the periapsis must be close to correct, assuming you're burning at apoapsis.

So the idea is to match the shape of the segment of the projected orbit before the encounter to the target orbit as perfectly as possible? I guess that would work if you can see at least most of one half of the orbit which should always be possible. So the verdict is it's comparing to the hypothetical orbit if there were no encounters for determining if it's correct?

You don't need to be at apo or pe to change the orbit. Try burning radial

That's not the problem. Yes I know you can match the orbit with a single burn if you can intersect it anywhere. The issue is how do you tell if you're in the right (hypothetical) orbit when you can't see the whole thing because there's inevitably an encounter coming up.

[Anquietas314]

So the idea is to match the shape of the segment of the projected orbit before the encounter to the target orbit as perfectly as possible? I guess that would work if you can see at least most of one half of the orbit which should always be possible. So the verdict is it's comparing to the hypothetical orbit if there were no encounters for determining if it's correct?

Well, yeah but you don't really need much of the orbit at all; 20% of the orbit should be good enough in most cases... maybe as low as 5% if it's almost perfect. As long as the curves match up very closely, the overall orbits won't be far off each other.