non-optimal ejection burns

[Fail-o-matic]

How do you launch to other planets in a non optimal window? Say you want to send a probe to duna on day 1 instead of day 50 or 60, same procedure or are the calculations different?

[RexKramer]

To reach another planet, you need to reach a point on the planet's orbit at roughly the same time the planet does. If you launch outside of an ideal launch window, you won't reach the point at the same time the planet does. There are a couple of crude solutions to this.

If your burn causes you to reach the target orbit ahead of the target planet, you can burn longer to arc beyond the target orbit, allowing the target planet to 'catch up' and intercept. (Assuming the target is outside of Kerbin..)

If you reach the orbit behind your target, you can adjust your solar orbit at AP to 'catch up' to the target planet, which will require at least one more orbit around the sun.

The same techniques can be used to reach inner planets. The penalty you pay for launching at non-ideal times is you use more DV, and it takes longer to reach the planet.

[kerbochuck]

If you don't mind mods, KER and protractor tell you the angles.

Essentially you need to burn at the right time so that you get to your target when it's in the right spot.

[Snark]

In terms of what you see when you look at the map window while you're planning your nodes:

If you launch in the optimal window (i.e. you're doing a Hohmann transfer), then you'll notice that your transfer orbit is tangent to both the origin and the destination. That is, it's an ellipse whose periapsis just touches the inner planet's orbit, and whose apoapsis just touches the outer planet's. The orbits don't "cross"-- they gradually converge, so that you're travelling parallel to the planets in question.

A non-optimal launch will involve violating that in some fashion. It's visible in the map window as a path that crosses the orbit of either the source or the destination or both.

Answering "how" isn't super trivial-- there's only one optimal trajectory, but there are an infinite number of non-optimal ones, depending on how much time you're willing to take.

I usually end up doing one of the following, depending on whether the target planet is ahead of the ideal position for launch window, or behind (and by how much).

If the target planet is ahead of the ideal position (but not too much):

I eject slightly inwards. Instead of launching in the same direction as my origin planet's orbit around the sun, I launch somewhat sunwards of that, so that my path takes a slight "short-cut". I aim to arrive near-parallel to the target.

If the target planet is ahead of the ideal position (by a lot, like a large fraction of 180 degrees):I go for a bi-elliptic transfer. I make retrograde burn to drop my orbit periapsis down close to the sun, then do a prograde burn at solar periapsis to zip out to the target planet. I aim to arrive near-parallel to the target.

If the target planet is behind the ideal position (but not too much):

I eject slightly radially outwards. That is, instead of launching strictly prograde in the direction of my origin planet's orbit, I eject somewhat outwards from that. I aim to arrive near-parallel to the target.

If the target planet is behind the ideal position (by a lot, like a large fraction of 180 degrees):

This is just a nasty situation, it's hard to come up with a solution that isn't incredibly wasteful of dV. This is basically the worst-possible situation, and if I find myself in this position, I generally just wait it out until it's something more manageable.

[satnet]

You might want to check out http://alexmoon.github.io/ksp/. It will calculate a wide range of launch windows. You can specify departure, time of flight, and transfer type. I use it all the time and find it invaluable. When you run it the output will show the ejection angle/delta-V for the most optimal transfer with your parameters it could find. You can also click anywhere on the plot to see the ejection angle/delta-V for a different transfer.

[quietghost]

My rule of thumb is that when it comes to orbital maneuvers, it is always better to burn late than early. So if you know what the angle between Duna and Kerbin should be from the tools mentioned above, a good strategy is to wait an extra day or 2 before actually ejecting. The added dV is minimal compared to adjusting the departure angle. Normally in life people say early is better than late, but in space late is always better than early

[Sharpy]

You might want to check out http://alexmoon.github.io/ksp/. It will calculate a wide range of launch windows. You can specify departure, time of flight, and transfer type. I use it all the time and find it invaluable. When you run it the output will show the ejection angle/delta-V for the most optimal transfer with your parameters it could find. You can also click anywhere on the plot to see the ejection angle/delta-V for a different transfer.

This - and PreciseNode. Using it you can create nodes for these planned burns and just observe how the trajectory changes.

For example, instead of waiting for optimal Eve transfer, I choose a slightly less dV-optimal trajectory that would take some half a year longer (but considering the optimal window was in more than half a year, that was a fair deal.) Now my Eve-bound probes are nearly skirting Duna orbit. Instead of fuel-optimal pro- or retrograde burn relative to the Sun, the ejection burn had a strong Radial component, which (depending on location) can simultaneously raise apoapsis and drop periapsis. The orbital injection is a neat tangent, like in Hohmann transfer, but the ejection from Kerbin system was "outwards", into a wide arc that gives Eve time to catch up with the ship that now is trailing behind Kerbin.

[Streetwind]

Bonus: you can get the Transfer Window Planner as a mod ingame, too!

[eddiew]

Bonus: you can get the Transfer Window Planner as a mod ingame, too!

I second this, it's an absolutely brilliant tool for planning a burn ahead of time, and it adapts to your current game conditions, including any planet packs you might have installed. MechJeb is also capable of giving you an in-game, clickable plot of when you can burn for how much fuel cost, and will set up the manoeuvre node for you.

If you really have to go ahead of the optimal window, then you're in for a lot of extra delta-v to get an encounter, arrival will be at a higher velocity than ideal, and will require a harsher aerobrake or more fuel to stop at the far end... overall you don't want to do travel out-of-window except for some form of challenge, or rescue mission with life support involved TWP is great for planning your missions around the natural cycle of planets ^^

[KerikBalm]

Use more dV... if its far away from launch, use a lot more dV....

Your capture burn will require a lot more dV too (unless you just do a flyby)

Like what Snark said... but what he said is only valid for Duna/Dres/Jool/Eeloo... its reversed if the planet is closer to the sun than Kerbin.

However, that is just Moho and Eve, and Moho's window comes up very frequently.

So really, its just a consideration for going to eve

[Sharpy]

Use more dV... if its far away from launch, use a lot more dV....

Your capture burn will require a lot more dV too (unless you just do a flyby)

There are also good transfer windows that require excessive time. "Optimal" is subjective.

Take this plot.

Eve, early 3rd year. You can wait until day 56 and begin a trek that will take almost two years for 2,741 m/s. Or you can launch today for 2795 m/s and be there in 620 days for a pitiful 54m/s more. Or you can dish out another thousand (3,666 m/s total to be precise) and shorten the journey to 371 days. Or if you're in a hurry, dish out some 6500m/s for an express 144 days.

[cicatrix]

There are no transfer windows, there is not enough dV. Basically, if you have enough dV you can launch your craft at whatever time you want.

I was reading an old sci-fi and there I found an 'over the Sun' interplanetary trajectory (entering the ~90 degree inclination orbit around the Sun and passing to the target planet). I'm wondering if it's possible to buid a craft with dV enough for that.

[Padishar]

Normally in life people say early is better than late, but in space late is always better than early

I assume you've never done a suicide burn...

[KerikBalm]

There are no transfer windows, there is not enough dV.

What?

Basically, if you have enough dV you can launch your craft at whatever time you want.

But "enough dV" varies with time, and there are intervals of time ("windows") where the required dV is significantly lower.

I was reading an old sci-fi and there I found an 'over the Sun' interplanetary trajectory (entering the ~90 degree inclination orbit around the Sun and passing to the target planet). I'm wondering if it's possible to buid a craft with dV enough for that.

its possible, yes, but extremely cost prohibitive and inefficient.

You can get going nearly as fast as you want, as long as you don't mind your rocketgrowing arbitrarily big, and having an aribtrarily small payload.

[cicatrix]

What?

But "enough dV" varies with time, and there are intervals of time ("windows") where the required dV is significantly lower.

its possible, yes, but extremely cost prohibitive and inefficient.

You can get going nearly as fast as you want, as long as you don't mind your rocketgrowing arbitrarily big, and having an aribtrarily small payload.

My whole message was oozing with sarcasm. Anyway, never mind.

[Starhawk]

There was a challenge a while back called the Fastest Small Step which was all about getting to another planet's surface as quickly as possible.

Those are some non-optimal ejection burns! Duna in seven days, Eve in six. Of course the delta-v costs are enormous. And this was done when aerobraking was infinitely magical.

Happy landings!

[Sharpy]

I assume you've never done a suicide burn...

That's "Early" too except upside down

Optimal landing is mostly the reversal of optimal launch; enter the orbit as early as possible, leave the orbit as late as possible. Perform ejection burn in as low orbit as you can, perform capture burn at as low periapsis as you can. Accelerate early, brake late.

I was reading the thread about K-drive. One of the variants that unfortunately was lost forever (...if it ever existed) was "Prograde Drive". It didn't provide any standard acceleration, it just turned your velocity vector to whatever direction you turned your ship, relative to current SOI. Entering Sun retrograde orbit? Exit Kerbin SOI, turn retrograde relative to Kerbin, you've just done 18500m/s worth of a maneuver turning your orbit from prograde to retrograde. Pure sci-fi, you can perform gravity assists with the same planet over and over as fast as you can cross it SOI or perform Oberth maneuver over any planet for as long as you desire, staying in its low orbit at indefinitely high speed simply by constantly aiming just above its horizon.

[Drethon]

When time is not an issue you can use partial Holman transfers (not sure that is a term).

I made it to Moho with relatively low delta v by first entering Kerbol orbit just outside of Kerbin's SOI and time warping to the nearest AN/DN with Moho. Then I drop the periapsis to intersect with Moho's orbit. Finally at each periapsis I drop the apoapsis down a bit to get the speed closer to Moho or I just make one large burn to intersect with Moho at the next periapsis.

This reduces the burn to orbit by quite a bit, possibly down to just the plain matching burn. Horribly game time inefficient but fairly low delta v without waiting for the right transfer.

[Sharpy]

When time is not an issue you can use partial Holman transfers (not sure that is a term).

I made it to Moho with relatively low delta v by first entering Kerbol orbit just outside of Kerbin's SOI and time warping to the nearest AN/DN with Moho. Then I drop the periapsis to intersect with Moho's orbit. Finally at each periapsis I drop the apoapsis down a bit to get the speed closer to Moho or I just make one large burn to intersect with Moho at the next periapsis.

This reduces the burn to orbit by quite a bit, possibly down to just the plain matching burn. Horribly game time inefficient but fairly low delta v without waiting for the right transfer.

I'd say you aren't saving anything as you're losing Oberth effect benefits of insertion burn from a high speed. What you're doing is essentially "spiralling down" which is considerably less efficient than Hohmann.

[Lighthawk]

When time is not an issue you can use partial Holman transfers (not sure that is a term)...

Huh...I've used that technique to simplify orbital rendezvous for docking, never considered trying it for planetary transfers though. I will say, whatever it lacks in efficiency compared to normal methods, the method is stupid easy to use and damn near impossible to screw up.

[wrexecute]

Bonus: you can get the Transfer Window Planner as a mod ingame, too!

This thing is awesome with Alarm Clock! The details it adds to alarms when using the plotter is just amazing. Even better, pair it all up with KCT and then you really gotta grind and design on a deadline especially with Duna coming up around the 225 day mark in a fresh career.

OP: While non-optimal ejections are totally possible and explained very well in some posts already, waiting for the right window has been made less uncertain and less boring with the help of Transfer Window Planner, Kerbal Alarm Clock, and Kerbal Construction Time.