Eyeballing Transfer Windows

[Streetwind]

I know there's plenty of websites and ingame tools that tell you the optimal times to depart Kerbin for interplanetary voyages. But I've always given myself the rule that before I let automation handle something for me, I must learn to do it by hand. So now I am looking at the solar system view and scratching my head.

On some planets, I cannot get a proper encounter no matter how hard I try; at least, not directly from Kerbin. I could go into a solar orbit and treat it like a rendezvous and it would probably work, but I assume that these planets are simply well out of their ideal transfer window right now.

On other planets I can get an encounter directly from Kerbin. The question here is, how do I judge if what I have plotted out now is a good course? How do I work out if waiting will improve or worsen the situation? This is something I've never thought about before.

Again, I'm not looking for the already well-known and calculated numbers. I'm looking for the means to calculate - or at least, eyeball - them myself. If I end up a whole month off because I use huge simplifications, that's fine. I just want to be roughly able to tell "bad time to launch" apart from "good time to launch", and "situation improving" from "situation worsening".

[Skorpychan]

Far as I know, it's about phase angles from kerbin. That should give you somewhere to start from. Wikipedia is a goldmine of useful maths for KSP.

[Kasuha]

First point: correct ejection from Kerbin system is always along the Kerbin's orbital path along Sun. When you're planning your manual transfers, always adjust your burn s you exit in such direction. Note that your ejection path from the maneuver will be curved. What counts is where it crosses the SOI and where that point is in the interplanetary spave (the map will show you the point if you put your mouse over the SOI exit point).

Second point: to visit planets on higher orbits, you eject Kerbin-prograde. To visit planets on lower orbits, you eject Kerbin-retrograde.

Now to the matter: you need to plan your maneuver so you exit Kerbin SOI the way described above, and your orbit just touches the orbit of your destination when viewing the map from above

After that, put another maneuver to the ascent/descent point and add normal burn to match the inclination. If you do it right, nearest approach marker will appear at the point where the two orbits touch.

The angular difference between the two markers is how much you need the angular difference between the two planets to change. Meaning, at the point when you are at the marker, you need the planet to be at the marker, too.

Eeloo and Moho may be more tricky due to their eccentricity and inclination. It may be good idea to iterate that in a few steps. Other planets are usually ok - you plan the maneuver, notice the angle, fast forward till the angle matches, and plan and execute transfer burn.

[BlazeFallow]

Fromt the Cheat Sheet in the wiki, you will find this handy chart which displays all of the ideal Delta V requirements for interplanetary travel, coupled with the ideal phase angles to launch from. Print that sucker out on a piece of thin paper (the 100% recycled stuff works great!), set Kerbol as the center of your map view, and you can hold it up to the screen and eyeball to your hearts content. Note: this is only for Kerbin departures and returns. Between other planets... well you'll have to do some maths to figure that one out. But here is a lovely tutorial piece to figure out the other planets.

[Streetwind]

Now to the matter: you need to plan your maneuver so you exit Kerbin SOI the way described above, and your orbit just touches the orbit of your destination when viewing the map from above

After that, put another maneuver to the ascent/descent point and add normal burn to match the inclination. If you do it right, nearest approach marker will appear at the point where the two orbits touch.

The angular difference between the two markers is how much you need the angular difference between the two planets to change. Meaning, at the point when you are at the marker, you need the planet to be at the marker, too.

Eeloo and Moho may be more tricky due to their eccentricity and inclination. It may be good idea to iterate that in a few steps. Other planets are usually ok - you plan the maneuver, notice the angle, fast forward till the angle matches, and plan and execute transfer burn.

That helped a lot! I managed to get a Moho intercept. My next question is now: is there a way to estimate if that is a "good" intercept or not? And: does it matter, or should I just go with whatever intercept I get since it doesn't matter much? I would assume that travel times would be greatly affected. The intercept I hit during my test was about 27 days.

Upon trying Eve, I found that the ascending/descending nodes I'd need for an inclination burn were a few degrees behind Kerbin and a few degrees ahead of periapsis, respectively. Which would mean I wouldn't be able to do the inclination burn until I was very close to already touching the target orbit. Is that a situation where I should recognize that the window is way off, or does that simply mean I need to maneuver differently?

Fromt the Cheat Sheet in the wiki, you will find this handy chart which displays all of the ideal Delta V requirements for interplanetary travel, coupled with the ideal phase angles to launch from.

Yeah, that's more or less exactly what I don't want, but thanks anyway

[problemecium]

I've wondered since I started playing what's wrong with just launching interplanetary probes whenever and simply maneuvering when the ship's phase angle lines up. Am I missing something?

[WhiteKnuckle]

I've wondered since I started playing what's wrong with just launching interplanetary probes whenever and simply maneuvering when the ship's phase angle lines up. Am I missing something?

You mean leaving them in a "parking orbit" around Kerbin and then just waiting until you get a launch window? Nothing wrong with that at all.

You can even do it with manned missions since in the stock game Kerbals do not need pesky things like food, water, or air to breath, so who cares if they spend 6 months in orbit waiting to leave.

[WhiteKnuckle]

Upon trying Eve, I found that the ascending/descending nodes I'd need for an inclination burn were a few degrees behind Kerbin and a few degrees ahead of periapsis, respectively. Which would mean I wouldn't be able to do the inclination burn until I was very close to already touching the target orbit. Is that a situation where I should recognize that the window is way off, or does that simply mean I need to maneuver differently?

You only need to do inclination burns at a node if you want your orbit to match the orbit of your target. You don't care about that, you just need either the ascending or descending node to fall at the same point as your encounter.

Since one of your nodes is near periapsis (which should be your encounter with Eve) you shouldn't have to move it much to get that node right on the encounter.

The only downside to not matching orbital inclination first is that you will not be in an equatorial orbit around Eve, you may not care about this, or if you do you can always adjust after you get into an Eve orbit (but it will cost more deltaV to do it this way)

Edited February 15, 2014 by WhiteKnuckle

[Streetwind]

Ah, I see.

Well, it wouldn't work anyway since Eve itself won't be anywhere near the point of contact, so the window is likely way off anyway Was just wondering if the relative positioning of the nodes had any significance.

[palioxis1248]

I don't really understand the obssession with transfer windows, but that's probably because I overengineer the deltaV capability of my rockets. Basically if you have really capable vehicles (i.e. nukes and the like), you don't really need to eject prograde and time it right. Leaving at a 10 degree angle from Kerbin usually tacks on another .5kps, but at least it gets you there without you having to wait in-game years for transfer windows.

My advice if you're trying to do this is to start in a circular orbit, and start with a regular prograde ejection. Then, once you've pulled out the opposite solar apsis until it touches the planetary orbit of choice, trying rotating the maneuver node around the circular orbit and adding some extra delta v and see where that gets you. Generally speaking if the planet is too far behind when you reach its orbit, don't change the ejection angle and just increase ejection delta v. Of course you should aerobrake wherever available. If the planet is too far in front, rotating the ejection angle towards the sun makes you cut across the inside of Kerbin's orbit, which should allow you to nail an intercept with the planet later on.

Not exactly anything to do with transfer windows, but I hope this helps!

[WhiteKnuckle]

I added an edit to my earlier post, and would like to further clarify that whenever possible you should do your inclination burn at a node, it's just not the end of the world if you don't.

Ah, I see.

Well, it wouldn't work anyway since Eve itself won't be anywhere near the point of contact, so the window is likely way off anyway Was just wondering if the relative positioning of the nodes had any significance.

Did you already launch the ship? In this case you should do the burn at the first node you come to, this way your orbit will match Eve's orbit and you may get an encounter further on in the orbit or a few orbits from now (see palioxis's post about fiddling with an orbit to get an encounter outside of a transfer window)

[Streetwind]

No, I've not launched anything yet, gotta design a spacecraft first. Also, from what I've estimated so far, I seem to be much too early on most encounters apart from Moho, so waiting a bit longer won't hurt.

[BlazeFallow]

As for the dislike of my post, I understand, it gives the precise values... But in fairness, you're asking how to "eyeball the transfer windows" which without knowing roughly where the ideal window is, isn't going to happen. The sense will not come out of just looking at the picture, because each planet moves at different speeds and orbits, so comparing one launch vector to another just doesn't work. So again, I say that unless you want to do the full maths (as i agve in the latter half of my post) then simply knowing roughly the phase angles of each planet's prime transfer windows will let you eyeball it. You don't HAVE to print it out, if you still want the guess work of "I guess Duna looks like it's 45 degrees ahead of Kerbin...."

[Streetwind]

Au contraire, it is happening quite well. I've plotted quite a few vectors and timewarped a bit in sandbox mode, and now I've got a rough idea what I need to watch out for (and how to force a transfer outside of a proper window). I'll freely admit that it'll be imprecise, but it certainly won't be random.

[Vanamonde]

Not an answer to the question but a word of advice: launch windows are not all that critical. There is a maximally efficient instant, but if you miss it, it's merely a matter of burning a bit more fuel, proportionally to the degree of the deviation from ideal. Just eyeball it as close as you can, and then play with the maneuver node until you get an intercept. If you're having trouble, get as close as you can and then plot a mid-course correction, which will often be enough to close the gap and achieve an intercept.

[metaphor]

My next question is now: is there a way to estimate if that is a "good" intercept or not? And: does it matter, or should I just go with whatever intercept I get since it doesn't matter much?

It does matter where you have your intercept, especially with Moho, less with other planets. If your transfer orbit crosses Moho's orbit at a large angle, you're going to have to do a big capture burn to get into Moho orbit. If your transfer orbit is almost tangent to Moho's orbit at the encounter, it's going to take a lot less delta-v to capture into orbit.

[MockKnizzle]

Quick rules of thumb for eyeballing:

- To transfer up to something higher, your target needs to be ahead of you. To transfer to something lower, it must be behind.

- For all but the closest bodies to Kerbin, the transfer windows are at nearly 90 degrees.

- Duna and Eve are about 45ish degrees ahead and behind, respectively.

- Moho sucks, eyeballing there can be dangerous since delta-V requirements are so high even with a perfect transfer. Any inefficiency can add nearly 3 or 4 km/s to your capture burn.

- Returning to Kerbin? Good luck, I have no advice.

[StrandedonEarth]

Kerbal Alarm Clock has a basic planetary transfer window calculator. Not terribly accurate IMO, but it's the only one I've used (and it's why I got KAC in the first place; now it's indispensible) and it's close enough for Kerbin government work.

[sgt_flyer]

I eyeballed at first, now i use alexmoon's very handy calculator, http://alexmoon.github.io/ksp/ which gives you on which day the next transfer window is wayyy more reliable than eyeballing it, if you want to do transfers without mods