How to go from Ike to Minmus?

[paul_c]

I have a vehicle which is in low orbit of Ike; and I want to get it to the surface of Minmus. According to this: https://13375.de/KSPDeltaVMap/ it should take ~1050m/s of dV. The vehicle has 2264m/s dV. But what's the route? Do I

1. do a burn to just leave the orbit of Ike and drop into a wide orbit of Duna (being careful to avoid another accidental encounter with Ike, its a big blob)
2. Wait/timewarp to the relevant transfer window, then upon the encounter with Kerbin, aim to enter an elliptical orbit (lets say, aim for 69km Pe, it doesn't have heatshields etc but I can retract the solar panels, then do a minimal retro burn to be captured in orbit by Kerbin) with an Ap beyond Minmus'
3. Then maybe a plane change
4. Wait or gently tweak things for an encounter with Minmus?
5. Minimise the Pe, enter Minmus orbit, then de-orbit, land etc?

[king of nowhere]

1) that's the best way to make long burns, because you take advantage of oberth effect. Duna to Kerbin, though, is only a small nudge, so you probably are better off not going for the assist

2) yes. aerobrake as much as you can, again, a duna-kerbin transfer shouldn't have a huge intercept. according to the launch planner, it would be 100 m/s to enter an elliptic orbit around kerbin

3) you can, but no real need. when minmus is generally in the right position, and you are in the high part of your ellipse, make a straight path to minmus. you will be going slowly in the high part of your orbit, so the manuever should be cheap.

your plan would still work fine, though

In general, you may even spend less than 1050.

[Streetwind]

The main issue is that you must do a Duna-Kerbin transfer, and that will come in towards Kerbin from a specific direction, and Minmus is unlikely to be in the place you need it to be based on that direction.

Expect to spend either significant extra time or significant extra dV once in Kerbin's SoI to make a Minmus encounter happen.

You can probably do a direct Duna-Minmus transfer, but the amount of dV that costs compared to aerobraking at Kerbin is likely prohibitive. It's also going to be very fiddly and trial-and-error.

[paul_c]

Made it with 22m/s dV left!

The burns were (approx):

1. 100 m/s to leave Ike for a wide Duna orbit
2. ~40m/s to adjust to not encounter Ike again
3. ~600m/s to do Duna > Kerbin. Due to my very wide Duna orbit, the planets didn't really align with my ejection angle, so I tried the 'first' one that looked about right - I might have been able to do a quicksave and try the 'next' orbit of Duna to see if it were better...but if it wasn't....
4. ~600m/s to insert into a Kerbin orbit, with an adjustment of inclination just as I entered its SOI to encounter Minmus on the way out

Then the problem was, I was heading for Minmus with way much more than normal velocity. So the retrograde burn to insert into Minmus orbit was about 650m/s, then the rest was used for the de-orbit and landing.

[Dr. Kerbal]

A simple guide from getting to Ike and Minmus for simple folk:

1. Go from Ike to Duna

2. Wait a for a Duna to Kerbin transfer window (same as Kerbin to Duna Transfer window) (45 degrees)

3. Burn for Kerbin.

4. Get to Kerbin.

5. Orbit Kerbin.

6. Get Encounter with Minmus.

7. Orbit Minmus.

8. Done

 

 

[king of nowhere]

3 hours ago, paul_c said:

Made it with 22m/s dV left!

The burns were (approx):

1. 100 m/s to leave Ike for a wide Duna orbit
2. ~40m/s to adjust to not encounter Ike again
3. ~600m/s to do Duna > Kerbin. Due to my very wide Duna orbit, the planets didn't really align with my ejection angle, so I tried the 'first' one that looked about right - I might have been able to do a quicksave and try the 'next' orbit of Duna to see if it were better...but if it wasn't....
4. ~600m/s to insert into a Kerbin orbit, with an adjustment of inclination just as I entered its SOI to encounter Minmus on the way out

Then the problem was, I was heading for Minmus with way much more than normal velocity. So the retrograde burn to insert into Minmus orbit was about 650m/s, then the rest was used for the de-orbit and landing.

you could have optimized much better

1-2 those burns made things a bit easier to plan, but they were absolutely not needed, and indeed they were even detrimental. you gain absolutely nothing on going from ike to high duna orbit, and in fact you lose some obert effect.

3 600 m/s looks just about right for a duna-kerbin intercept

4 huh? this looks completely wrong.  you should need very few deltaV to get captured in orbit around kerbin, unless you were coming in from a bad transfer manuever.

5 again, this looks way too much.

i am reluctant to try it because the kerbalism mod makes a simple test like that much harder, but maybe i'll give it a try

[paul_c]

I have literally no idea how I would have planned Ike - Kerbin directly.....

I dare say it could be done more efficiently but I don't have any planning tool, just the "theory".

1 hour ago, Dr. Kerbal said:

A simple guide from getting to Ike and Minmus for simple folk:

1. Go from Ike to Duna

2. Wait a for a Duna to Kerbin transfer window (same as Kerbin to Duna Transfer window) (45 degrees)

3. Burn for Kerbin.

4. Get to Kerbin.

5. Orbit Kerbin.

6. Get Encounter with Minmus.

7. Orbit Minmus.

8. Done

 

 

This in theory would work but the problem is, burning to orbit Kerbin then burning to encounter Minmus is going to be potentially quite a lot; and kinda in the opposite sense (you'll be burning retrograde to be captured by Kerbin, then prograde to encounter Minmus) so I think this is the critical phase.

Duna-Kerbin transfer window is "Kerbin 75deg behind Duna"?? https://ksp.olex.biz/

[king of nowhere]

 

4 hours ago, paul_c said:

Made it with 22m/s dV left!

The burns were (approx):

1. 100 m/s to leave Ike for a wide Duna orbit
2. ~40m/s to adjust to not encounter Ike again
3. ~600m/s to do Duna > Kerbin. Due to my very wide Duna orbit, the planets didn't really align with my ejection angle, so I tried the 'first' one that looked about right - I might have been able to do a quicksave and try the 'next' orbit of Duna to see if it were better...but if it wasn't....
4. ~600m/s to insert into a Kerbin orbit, with an adjustment of inclination just as I entered its SOI to encounter Minmus on the way out

Then the problem was, I was heading for Minmus with way much more than normal velocity. So the retrograde burn to insert into Minmus orbit was about 650m/s, then the rest was used for the de-orbit and landing.

ok, i did a quick trial. and by quick, i really mean "i can't be bothered to do things properly", and i spent much more than i could have. but i still needed 500 m/s less than your trajectory

this is my "probe", very simple. I deactivated commnet to not bother about antennas and planetary occultations. the important part is the vacuum deltaV, 3643 m, to judge how much i spent

start: i cheated the probe in ike orbit. from there, without too much hassle i find an intercept with kerbin for less than 700 m/s. that's cheaper than going to duna orbit first.

Do notice that it's year 1, day 259: i am well outside of an optimal transfer window. the graph underneath shows that i could save 500 m/s launching at a more favorable time. but i didn't want to wait two years; i had this old save that was mostly empty and had the right time, so i used it.

Anyway, I overdid the burn. now i will need a course correction

I am waiting until i am in kerbin's SoI to make the course correction, even though it would be better to make course corrections earlier. I spend 60 m/s for it, when i could have gotten away with less than 10. but it was easier doing it this way. I also fix my orbital plane to come in close to minmus plane. This far from a planet, plane changes should really be cheap

also notice my periapsis, 45 km. i eyeballed this as a good altitude for aerobraking. my intercept speed is 500 m/s, i will need to lose some of it. remember, if i had launched in a proper transfer window, i wouldn't have to spend so much for capture

At periapsis I almost burned up, which means my periapsis was quite good. I aerobraked some 100 m/s, i need to use the rocket to provide the rest

400 m/s of additional burn see me in an elliptic orbit

apoapsis is quite right for minmus intercept. i just need a small nudge to get away from kerbin's atmosphere

now, since i've done a good job picking my entry trajectory, i only need a very small burn to fix my plane. though you can get to minmus with a wrong orbital plane if needed.

minmus is on the opposite side of the orbit, i just need to wait some orbits to catch up

here i come close. actually i am a bit late, which is bad. I can delay my craft easily by raising my apoapsis, but i cannot speed it up. now, the proper thing to do would be wait another minmus orbit until i can get a better intercept. or go back one orbit, lower my apoapsis (when apoapsis is so high, you can easily increase/decrease your orbital time by one day with a few m/s), and increase it the next orbit. I didn't want to be bothered, but i could have saved fuel there too

So I make a radial burn. You normally want to avoid them, cause they are expensive, but when your orbit is so high and slow, they become quite cheap, and they are convenient ways to grab moons in eccentric orbits. 100 m/s for intercept, 250 for capture. then it's a normal landing

landed with 1943 m/s. i started with 3643, so i spend 1700 m/s exactly.

and that was on a hasty attempt, outside of an optimal launch window, and making course corrections in a very suboptimal way. I bet it would be possible to do it with less than 1000 m/s for a challenge.

I really have no idea what you did wrong to get such a high consumption

[paul_c]

3 minutes ago, king of nowhere said:

 

 

ly have no idea what you did wrong to get such a high consumption

I know what I did wrong - I forgot to do one of the correction burns far out from Kerbin. But since I didn't have any intermediate saves, etc, I just carried on and still made it there.

[OHara]

22 hours ago, paul_c said:

According to this: https://13375.de/KSPDeltaVMap/ it should take ~1050m/s of dV.   [...]
 But what's the route?

In very low contrast, as is popular with the 13375, there is a 'details' link that gives the route:

 To an intercept trajectory of Ike.      198 m/s
 To an elliptical orbit around Duna.      33 m/s
 To an intercept trajectory of Duna.     275 m/s
 To the orbital plane of Kerbin.           0 m/s
 To the edge of the influence of Kerbin. 143 m/s
 To the orbital plane of Minmus.           0 m/s
 To an intercept trajectory of Minmus.    22 m/s
 To a low orbit around Minmus.           176 m/s
 To the surface of Minmus.               198 m/s

The JavaScript on that page sums the delta-V for all the legs of the journey as given on the Community Delta-V Map (link) plus a 10% margin, configurable in its settings.   (That JavaScript also considers some direct legs from planet to planet, without going back to Kerbin in between, that are not in the community map.)

To make it on the given budget, you would have to perform each leg as efficiently as you would if you were doing it alone. 

That means waiting until Ike is in just the right position so that the burn "to an elliptical orbit around Duna" puts that elliptical orbit at just the right orientation so that when the Duna-Kerbin transfer window arrives, just a small additional burn at periapsis ejects you from Duna at the right angle to reach Kerbin.    Then as @Streetwind said you aerobrake a little bit at Kerbin to bring your apoapsis to the height of Minmus, and wait until Minmus passes by your orbit.

This amount extreme planning, using the tools of KSP would need a number of marker craft, cheated into various orbits, or some other creative solution.

The code on the webpage is good for quickly doing the sums, and interesting path-finding code to look at,  but I think the delta-V map is much easier to use for planning.

Edited January 5, 2021 by OHara

[Snark]

On 1/4/2021 at 4:16 PM, paul_c said:

1. do a burn to just leave the orbit of Ike and drop into a wide orbit of Duna (being careful to avoid another accidental encounter with Ike, its a big blob)

Not a good idea.  This would lose all benefit of Oberth effect. 

Basically, you would either want to stay in low Ike orbit and eject directly from low Ike orbit to Kerbin intercept with a single burn, or (possibly) eject from Ike orbit in Duna-retrograde direction to drop your Duna Pe down very low, and then do a prograde burn at low Pe over Duna.

Which one of those two would be better in terms of dV depends on the relative masses of the bodies in question and the height of the moon's orbit.  I'd have to crunch math to come up with the answer, but off-hand my instinct would be to keep it simple and go for a direct ejection burn from low Ike orbit.

(Naturally, when I say "direct ejection from Ike orbit", I mean that you'd eject in the same direction that Ike is traveling, so that your overall ejection speed from the Duna system would include Ike's orbital velocity around Duna, plus your orbital velocity around Ike, plus your ejection burn-- all three components pointing in the same direction.)

In general, you want to avoid doing large burns while in high orbit around anything.  It's very dV-inefficient.

 

For your arrival in Kerbin system:  the exact sequence of things will depend somewhat on your approach geometry, due to the inclination of Minmus' orbit.  There are several possibilities.  However, whichever specific approach you use, the first thing should be that your approach to Kerbin should have you on a trajectory that takes you to a very low Pe over Kerbin, and you do a burn when you're at Kerbin periapsis in order to capture to Kerbin and reduce your Ap down to about the altitude of Minmus.

(In other words, don't have your approach try to go directly to Minmus and try to slow down there from interplanetary speeds.  Do the capture burn in low Kerbin orbit.)

[king of nowhere]

19 hours ago, Snark said:

In general, you want to avoid doing large burns while in high orbit around anything.  It's very dV-inefficient.

unless you are making large changes in the angle of your trajectory, in which case high orbit is the only place to make them at a reasonable cost.

[Snark]

4 hours ago, king of nowhere said:

unless you are making large changes in the angle of your trajectory, in which case high orbit is the only place to make them at a reasonable cost.

Statement still stands.  Large burns at high altitude are inefficient.

Not sure what exactly you mean by "the angle of your trajectory"... do you mean your inclination?  If so, then yes, it's better to make an inclination at high altitude... because then it's not a large burn.

Note that I didn't say that all low-altitude large burns are efficient-- just that high-altitude large burns are.

There are certainly cases (such as a large plane change) where it's more efficient to a high-altitude small burn instead of a low-altitude large burn, precisely because it's small.

[king of nowhere]

4 minutes ago, Snark said:

Statement still stands.  Large burns at high altitude are inefficient.

Not sure what exactly you mean by "the angle of your trajectory"... do you mean your inclination?  If so, then yes, it's better to make an inclination at high altitude... because then it's not a large burn.

Note that I didn't say that all low-altitude large burns are efficient-- just that high-altitude large burns are.

There are certainly cases (such as a large plane change) where it's more efficient to a high-altitude small burn instead of a low-altitude large burn, precisely because it's small.

all you say is true, of course. but no, what i'm talking about is burns such as those i used to reach gilly

part of that is plane change, but not all. i'm also changing the way the elliptic orbit is pointing. I don't know the proper names, i said the angle of your trajectory because in those burns, the ship is going in one direction, and that direction is changed. and they are 100-200 m/s, which is not particularly small

perhaps the best way to describe them would be radial/antiradial burns. though those trajectories make liberal use of all three directions, but ultimately it's the normal/antinormal and radial/antiradial that get benefitted from happening in high orbit, while the prograde/retrograde component is basically to ensure that I intercept gilly at the right time.

radial/antiradial manuevers, like plane changes, are cheaper when the ship is slower. they are, however, rarely mentioned, because they are rarely useful and rarely used. they are, however, excellent ways to intercept a moon in an irregular orbit when your own starting elliptic orbit is not pointing in the right direction.

[Snark]

25 minutes ago, king of nowhere said:

<various good and well-explained stuff>

Sure, no argument there.  Again, though-- key argument is large burns.  Which is admittedly a relative term, but I'm describing burns that are large in relation to orbital-velocity-at-low-altitude.  For example, the maneuvers you've shown there are very small ones-- only 161.4 m/s and 120.2 m/s.  They also have the property that they're being done where they're being done because they have to be there, due to geometry.  (For example, if a ship is in a highly eccentric orbit and wants to raise its Pe, then the burn to do so has to be done at high altitude-- you can't raise your Pe higher than the altitude you're at when doing a burn.)

I'm not saying "never do burns at high altitude".  Some situations (like raising Pe) you have to, due to geometry.  Other situations (like plane changes), it's better to do at high altitude because you have the option of "large burn low, or small burn high" and the small burn is better.

What I'm talking about is neither of those cases-- I'm talking about "large burn low" versus "large burn high", such as arriving at Minmus from an interplanetary trajectory.  In such cases, doing the big burn is better done low than high; it's more efficient to do it at low altitude due to Oberth effect.  Oberth is your friend.  In the case of someone arriving at Kerbin on an interplanetary trajectory and wanting to go to Minmus, in general it'll be better if they arrive at low Kerbin orbit and do a big braking burn at low altitude, followed by some very small burns up at Minmus altitude, instead of trying to go to Minmus directly.

[king of nowhere]

15 minutes ago, Snark said:

I'm not saying "never do burns at high altitude".  Some situations (like raising Pe) you have to, due to geometry.  Other situations (like plane changes), it's better to do at high altitude because you have the option of "large burn low, or small burn high" and the small burn is better.

i know. i wasn't disagreeing with you. i was merely expanding a bit, in that there are some relatively expensive burns that are actually convenient in high orbit in some relatively rare situations.

 

by the way, i didn't have to make those burns there. the "standard" approach to reach a target in a different orbit would have been to match the plane, then change your orbit to intersect the target in one point, then keep burn at the intercept to change your orbital time and set up a rendez-vous in a future orbit. it's how the tutorial teches you. but in that situation, both nodes are very close to eve, resulting in a very expensive manuever. and then there's gilly's orbit being eccentric, and my orbit being eccentric with another angle, and this resulting in a high intercept speed. the radial manuever is a shortcut to go from an elliptic orbit to another, very different elliptic orbit without paying the crazy costs normally associated with that.

[paul_c]

I'm finding the same problems with a Laythe trip. Am I right in thinking, as a general principle:

* If you encounter a planet then (assuming you do a burn close to Pe, to get into a highly eccentric orbit just within ite SOI), if you have an encounter with a moon on the way out from the planet, you need to arrive behind that moon, as close as possible to it, to get the maximum gravitational assist but that will also go the right way ie slow you down and reduce the main planet's orbit?

* You can "choose which side" of a moon you encounter, making it a helpful/unhelpful in the grand scheme of what you're trying to do.

* It can also mess up your inclination (or improve it....but KSP doesn't show this numerically, only the graphic of the orbit)

* As well as any control you have on the effect of the encounter, you might just have a situation where the phase of the moon isn't right, so its gravitational effect is undesirable/best avoided?

I got into a situation where I encountered Laythe on the way back. My speed was about 4000m/s at Laythe Pe. Any retrograde burn resulted in splatting into Jool. So I was forced to burn about 1700 m/s to be fully captured but that meant I had enough to descend to Laythe, but not get back into its orbit. Any attempts to go below 50km at that speed resulted either in a fiery death or minimal effect (eg 49km didn't do much; 48km burnt me up). 

[OHara]

14 minutes ago, paul_c said:

I'm finding the same problems with a Laythe trip. Am I right in thinking,
[ . . . ]
I got into a situation where I encountered Laythe on the way back

Could you give more context about what you are trying to do ?  I'm not sure if you are thinking about the return journey, or slowing down when reaching the Jool system.

16 minutes ago, paul_c said:

if you have an encounter with a moon on the way out from the planet, you need to arrive behind that moon, as close as possible to it, to [ . . . ] slow you down and reduce the main planet's orbit?

This sounds backwards to me (unless I misunderstood your goal).  If you want to slow down your orbit around Jool--- and if your are orbiting Jool in the same direction as that moon is --- you can do that by passing in front of a moon.  Then the moon turns your path backwards a bit, relative to your orbital direction around Jool, resulting in your going slower relative to Jool.

Forum conversations are really good here, but also, trying out these gravity assists in KSP is one of best qualities of the game, so I recommend playing around a lot in-game.

[paul_c]

I have successfully made a spaceship which splashed down at Laythe, but in doing so I used all its fuel up. So it and 3 kerbals are all floating. It was quite big, with loads of science and a science lab so its done most of its job (there's a massive bonus in running the experiments in a science lab compared to transmitting the results back, or even retrieving them).

Now the challenge is to get them back. I thought I would split it up by sending a small ship which can land nearby, the Kerbals swim over to it, then it can return to Laythe orbit. Then later, I will send some other ship, park it in nearby in Laythe orbit (never going down to the surface then up) which can get back to Kerbin.

I built a big thing with lots of deltaV, but it took a lot to get it to Laythe. It was all a bit random. It was very far away from the Kerbals (that's another topic entirely!) and had about 2300m/s dV left. I tried to get it back to orbit but it didn't do it...I guess I need about 3500m/s for Laythe surface-orbit. 

I'll launch the (nearly) same thing again, this time I'll play around a lot more with moon (there's loads to choose from...) encounters to get the Jool orbit sensible and then just do loads of rotations to get a decent Laythe encounter etc.....

[king of nowhere]

2 hours ago, paul_c said:

I'm finding the same problems with a Laythe trip. Am I right in thinking, as a general principle:

* If you encounter a planet then (assuming you do a burn close to Pe, to get into a highly eccentric orbit just within ite SOI), if you have an encounter with a moon on the way out from the planet, you need to arrive behind that moon, as close as possible to it, to get the maximum gravitational assist but that will also go the right way ie slow you down and reduce the main planet's orbit?

* You can "choose which side" of a moon you encounter, making it a helpful/unhelpful in the grand scheme of what you're trying to do.

* It can also mess up your inclination (or improve it....but KSP doesn't show this numerically, only the graphic of the orbit)

* As well as any control you have on the effect of the encounter, you might just have a situation where the phase of the moon isn't right, so its gravitational effect is undesirable/best avoided?

I got into a situation where I encountered Laythe on the way back. My speed was about 4000m/s at Laythe Pe. Any retrograde burn resulted in splatting into Jool. So I was forced to burn about 1700 m/s to be fully captured but that meant I had enough to descend to Laythe, but not get back into its orbit. Any attempts to go below 50km at that speed resulted either in a fiery death or minimal effect (eg 49km didn't do much; 48km burnt me up). 

it seems you are trying to dance the gravity assists among the moons of jool. that can be very rewarding indeed, or it can also lead to suicide thought.

but as a general rule of thumb, if you pass in front of a planet, you get slowed down, and if you pass behind, you get accelerated. this is the basics of gravity assist. so you don't need any expensive manuevers in reaching jool. pass in front of laythe duing your jool approach, and you'll be in jool orbit for free. however, you need to keep a high periapsis above jool when you enter, because this manuever will also lower your periapsis. indeed, a gravity assist will basically give you a prograde/retrograde burn (together with other components, but those are less easily predicted), and since you are neither at periapsis nor at apoapsis, the burn will lower/raise both your periapsis and apoapsis.

now you are in elliptic orbit around jool. from there, in my experience it is better to head for tylo. you can reach there with 200 m/s if you get the assists right. doing anything else is way more complicated; in theory you can use tylo to raise periapsis and laythe to lower apoapsis, but i never did much successfully without spending more in correction burns than i would gain otherwise. after you get captured into an elliptic tylo orbit, if you wait the right time you can get to laythe with a few hundred m/s anyway.

so i suggest: jool periapsis - gravity assist by laythe - possibly another assist by laythe to get the course right - captured by tylo - wait in elliptic orbit around tylo - transfer to laythe

[ajburges]

Are you looking for an easy or efficient transfer? Burning to get deeper into Duna's gravity well is probably not efficient. The Oberth gains would need to outweigh the losses from going the wrong way. Given that Kebin transfer from Duna elliptical is ~380 dV there isn't much room there for Oberth savings.

Easy has already been covered:

• Leave Ike orbit for greater Duna orbit
• Transfer to Kerbin
• Enter Minmus orbit

Efficient relies on not spending dV without getting useful potential/kinetic energy

• Calculate the necessary dV to transfer to Kerbin from Ike orbit (dV_duna_escape)
• Calculate the dV burn necessary to have approximately dV_duna_escape on Ike -> Duna SoI change
• Time Ike exit burn such that Ike is approximately in the correct position for transfer burn (a few hours early so SoI change occurs at correct phase) and you burn on the correct trajectory for ejection. If you track at least 3 SoI changes with trajectory planning, you should get transitions that lead to a Hohmann like transfer from Duna to Kerbin and be near rendezvous.
  • You can skip to here with node planning, but mentioning the previous steps should help you visualize what you are doing.
• Halfway through the transfer, use fine trajectory adjustments to get a Munar intercept for an assist to capture into an elliptical kerbin orbit. Remember prograde, reduces energy, retrograde adds energy (or vice versa check against trajectory). You can also aerocapture, but a Munar assist capture will result in a higher Pe. Higher Pe is more energy in our orbit, The more energy in you orbit, the less dV you need for Minmus capture!
  • normal for plane adjustment
  • prograde + radial for pe height and time (Mun location) adjustments
• Munnar assist should help with Kerbin capture, adjust orbit for Minmus intercept.

[theAstrogoth]

Thought I'd add a link here that I saw in a related reddit discussion. It goes over the efficiency of interplanetary transfers when you're starting from either the Mun or Minmus and using either direct transfers from a moon or an Oberth maneuver (dropping into a highly elliptical orbit, skimming close to the planet before burning at periapsis).  All of the reasoning and math in that link should apply to going from Ike to Minmus, or any other moon/planet combo.

It looks like usually Oberth maneuvers are more efficient, but not always. The Oberth maneuvers are kind of hard to plan (at least for me), but I described an approach here that I think works with some tinkering.  The trick is in setting up the proper ejection angles/times.