How are you meant to figure out transfer windows in stock?

[InfernoSD]

I learned early on that the most efficient time to burn toward the Mun is when it passes over my prograde marker in LKO. I then learned that the same timing applied to Minmus, and the same timing also applies to when Duna passes in front of Earth's prograde. I found that to get to Eve I should wait until Earth was in front of what would be Eve's prograde.

This took some amount of trial and error to figure out, but I was quite proud of myself for doing so. I took it to be a principle of orbital mechanics and used it for hundreds of hours after. It served me well. I did have some difficulties lining up intercepts for more distant planets, but I assumed that was because my timing was off in some way, like because of the elliptical and inclined orbits some planets have.

That was until I tried timing a launch for Moho last week and failed repeatedly to get an intercept anywhere nearby. Every other planet had at least skewed close to this imagined relationship I had between planets, where to reach a planet in higher orbit I waited until it was "in front of" me, and to reach a planet in lower orbit I waited until it was "behind" me. Moho defies this relationship by being 108 degrees "in front of" Kerbin despite being in a much lower orbit. I had to look this up online and reschedule my launch.

Now that I know what the actual transfer windows look like, I'm left wondering. Is this principle I came up with just a coincidence resulting from how far apart these bodies are from Kerbin? Or was I nearing on some actual method?

Or could there be any other way to eyeball transfer windows? Mind you I play on PS4, so TWP was never an option anyway.

And does the real solar system exhibit similar launch windows to Mars and Venus?

[VoidSquid]

You can check the transfer windows for stock KSP here:

https://alexmoon.github.io/ksp/

26 minutes ago, InfernoSD said:

And does the real solar system exhibit similar launch windows to Mars and Venus?

Absolutely, yes:

https://mars.nasa.gov/spotlight/porkchopAll.html

 

In general, there are different methods to calculate the optimal time for a transfer, depending if you are orbiting the same CB (e.g. a vessel going from LKO to Mun, both, the vessel and Mun, are orbiting Kerbin) or not (e.g. a vessel going from LKO to Duna, one is orbiting Kerbin (which is orbiting Sun), while the target is orbiting the Sun itself).

Edited December 27, 2020 by VoidSquid

[king of nowhere]

I think you're supposed to mostly eyeball them. in most cases, they are large enough that launching a bit earlier or later won't make a big difference. that's why they are called transfer windows and not transfer pinholes. moho is a bit of an exception because it has a fast, eccentric, inclined orbit.

or perhaps you are meant to look online at the various tools available. or to use mechjeb.

[VoidSquid]

Just now, king of nowhere said:

or to use mechjeb.

OP said:

35 minutes ago, InfernoSD said:

Mind you I play on PS4, so TWP was never an option anyway.

 

[king of nowhere]

4 minutes ago, VoidSquid said:

OP said:

 

that would be very explicative, if I had any idea what is twp.

Anyway, i didn't say that the OP should use that. I said that perhaps you were supposed to. I suggested multiple ways, more or less accurate, in which one can find a transfer window. Perhaps nobody bothered to implement some calculator in the stock game because they said, there are so many methods available, those that reach that far will find something

[VoidSquid]

Just now, king of nowhere said:

twp

I think this mod is meant

 

[Incarnation of Chaos]

If i remember correctly, transfer windows occur at a angle between planets. There's tables with these for KSP in numerous places, and you could measure them with a protractor.

 

[king of nowhere]

3 minutes ago, VoidSquid said:

I think this mod is meant

 

oh, i see.

But no, no, no need to install any mod! I didn't even knew there was a dedicated mod (though, if i ever thought about it, i'd have guessed yes, because this game has a mod for everything)

One can always open the transfer window planner tool on the internet browser. which, if you're playing from play station, you can do on your mobile phone.

Granted, that bunch of data about ejection inclination and angle isn't very easy to replicate. but it at least gives some good indication on when you should make your launch, and what kind of deltaV cost you can expect. and you already know more or less in which direction to launch, so it's all about fiddling afterwards

[Spricigo]

On 12/27/2020 at 9:03 AM, InfernoSD said:

I took it to be a principle of orbital mechanics and used it for hundreds of hours after.

Well...No, it's not a principle just a perk of those celestial bodies. Maybe it is a design choice and maybe that is because it make things easier for the player once he notices this perk. In any case, 

Answering your question, for most things in stock we are supposed to "learn by doing". We go out, try thing, fail many times and eventually stumble upon something that works. However, the in game orbit happens to follow real word physics laws. Laws old enough* to be in the "should learn at school" category rather than "recent scientific breakthrough"

Of course, the phrase "solve for x" don't ring as "fun times are coming" for that many people. But there is quite a lot of useful information out there if you are willing to look at. If you are just looking for tools that makes the calculations for you, that is also available (some specific for KSP,  some are generic enough to be abe to setup for KSP). 

In any case, if you ask there, we will try to provide a good answer for you. 

*We can't even say it is simple newtonian physics because it is older than Isaac Newton.

[bewing]

In stock, if you have any craft in low Kerbin orbit, you can select it and create a maneuver node for it. Let's call that craft your "test craft" and say you want to go to Duna.

When you decide that you have enough tech and resources to go, then you "fly" your test craft from the Tracking Station. You select Duna as your target. You create a pure prograde maneuver node with an Ap just high enough to get to Duna's orbit. You notice that the "closest approach" marker for Duna shows Duna as being ahead of your Ap. For CBs that are in larger orbits than Kerbin, this means that you are too early for your launch. By judging how much of an angle, based on how big a fraction it is of a Duna year, you can calculate approximately how many days it will be until the transfer window.

You can do this a few times, several days apart, to narrow down your estimate.

Once you can create a pure prograde maneuver node for your test craft that has an Ap just above Duna's orbit, and the Closest Approach marker shows that Duna will be at your Ap at the same time as you are -- then you launch your real craft, create the real maneuver node, and do the real transfer. If you need some inclination to your burn, you can even launch your actual craft at sunrise or sunset into an inclined orbit.

On 12/27/2020 at 4:03 AM, InfernoSD said:

Or could there be any other way to eyeball transfer windows?

But the point is that the method that is embedded into the game for determining the timing of transfer windows is to create "hypothetical maneuver nodes" using test craft in orbit around your CB of origin.

And once you know when the transfer windows are, you can write them down -- because the planets are always in the same positions every time you start a new game.

Edited December 28, 2020 by bewing

[InfernoSD]

Thanks for the answers, people. I guess what bugs me about this problem is the amount of time involved in solving it. Designing rockets, you get instant feedback on how it flies when you go to the launchpad. Rendezvous and hohmann transfers, you can use manuever modes or fast forward for a moment to see what happens. Even the effects of a gravity assist can be found pretty easily by luck or by playing with manuever nodes for a few minutes. But if you aren't aware of the concept of transfer windows, as I wasn't, then you won't even know you're doing something wrong when you blast your way out of Kerbin's SOI. Learning what transfer windows are and when they are is something you can only achieve in-game by enormous fast forwards or by making many interplanetary missions.

Anyway, I decided to look into this a little deeper. I drew up this graph showing the relationship between the transfer window phase angle and SMA. This one depicts the Kerbol system but the relationship is the same for the Kerbin system or Jool system or any other.

From this, I guess you *can* easily eyeball some transfer windows. Going to a higher orbit is easy since the phase angle is within a narrow range between 0 degrees up to a maximum of 116.36 degrees. Going to a lower orbit is more of a challenge because the target can orbit its parent body once or multiple times during the transfer. The phase angle loops in the same way, hitting 180 degrees when the target has an SMA 46% of your starting orbit. Best thing to do is test to figure it out once and then write it down for future use.

I was going to take an extra step to chart out the deviance of prograde marker to target marker on the compass against SMA ratio. This would prove the point I was making in the OP, but the formula is a bit too complicated. From what I can tell, going to a higher orbit *does* always put the compass markers close together. In two very close orbits, the target marker is around 20 degrees above the prograde marker. In two very distant orbits, the target marker is around 20 degrees *below* the prograde marker. So it's really quite close. As for going to a lower orbit, I won't even touch that. It obviously goes crazy pretty fast below Eve.

[bewing]

3 hours ago, InfernoSD said:

As for going to a lower orbit, I won't even touch that. It obviously goes crazy pretty fast below Eve.

For lower orbits, you don't try to do it all at once. You leave Kerbin at the node for your destination and set your Pe at your destination orbit. When you get there, you set your Ap to create an intercept.  You'll get there in less than a year no matter how badly you time your transfer.

[HebaruSan]

4 hours ago, InfernoSD said:

I guess what bugs me about this problem is the amount of time involved in solving it. Designing rockets, you get instant feedback on how it flies when you go to the launchpad. Rendezvous and hohmann transfers, you can use manuever modes or fast forward for a moment to see what happens. Even the effects of a gravity assist can be found pretty easily by luck or by playing with manuever nodes for a few minutes. But if you aren't aware of the concept of transfer windows, as I wasn't, then you won't even know you're doing something wrong when you blast your way out of Kerbin's SOI. Learning what transfer windows are and when they are is something you can only achieve in-game by enormous fast forwards or by making many interplanetary missions.

Yup, precisely calculating a transfer is one of the more challenging tasks in the stock game. I would argue there is a bit of UI missing (quite understandable since it's a hard UI problem, NASA probably doesn't have a solution for it yet); to maintain the "learn by doing" approach, I don't think a list-of-transfers like my Astrogator mod should be in stock, but there could be more built-in support for @bewing's test craft method, without having to launch a test craft:

On 12/28/2020 at 4:37 PM, bewing said:

But the point is that the method that is embedded into the game for determining the timing of transfer windows is to create "hypothetical maneuver nodes" using test craft in orbit around your CB of origin.

Imagine if you could go to map view and click a special new button to create a "meta maneuver node" on your parent body's orbit rather than on your own craft's orbit. You would drag its handles and slide it around till you've pinpointed when the parent body will be in the right spot, and leave the "meta maneuver node" there as a marker. This would be similar to going to Mun or Minmus, and I don't think it gives the player any info they shouldn't have. Zooming back in to the parent body's SOI to plan your departure, the time, place, and speed when you should be hurtling away from Kerbin would be highlighted in the map view so you can make a regular maneuver node to hit it. I think this would provide a nice set of cues to get folks moving in the right direction with less outside help (whether from tools or other users).

[VoidSquid]

Ok, my 2 cents then (and what can you buy with 2 cents nowadays...)...

Preface: I'm playing KSP for 4000+ hours now. And I'm an Aspie. 

That said:

I totally hate how KSP stock is NOT giving me data, in particular when starting a new career. Even when having completed the tech tree, the stock data is scarce at best.

Mankind did send the first craft in orbit in 1957. At that time, the math and physics of the celestial mechanics of our solar system were well understood and known for more than 300 (!) years. For all major CBs we could calculate their orbit precisely. And KSP tells me to start my first rocket without any of those numbers/data? Seriously?

After, I assume at least 1000 rocket starts, I very much don't like to be bothered with manually steering to reach orbit, over and over and over again. After sending at least a couple of 100s vessel to Kerbin's moons as well as other CBs, I very much don't like to be bothered to manually find and optimize and execute a transfer node over and over and over again. It simply sucks for me nowadays.

Look at how air crafts or space crafts are controlled, and this is true for Sputnik, Apollo, SpaceX, Rosetta, ...: every maneuver is calculated and optimized in advance, the pilot (if there is one) is the one how oversees and controls things and who can at every moment (ideally) intervene and correct things and/or do them manually.

In the end, that's why I let MJ calculate and most often execute nodes, including prokchop transfers to other planets. Yes, I did it more than often enough for my own taste (and still have to if MJ fails). No, I don't feel it's cheating at all (which, for a single player game, is a ridiculous thing by itself). And I prefer doing critical stuff manually myself, like precisely landing a new, huge, heavy craft on another CB. Everything Eve: do it manually.

Personal note (Aspie here, mind you  ): it took me more than a year to finally find the courage to send some of my Kerbals outside Kerbin's SOI and land on another planet. It was KER, giving me the data, that made this possible.

Edited January 4, 2021 by VoidSquid

[vv3k70r]

On 12/27/2020 at 1:03 PM, InfernoSD said:

I found that to get to Eve I should wait until Earth was in front of what would be Eve's prograde.

You fly to da Mun, You fly to the Duna, You are breaking to Eve so Eve can catch You. Se relative speed to Sun for planets and moons against those reference points.

So You are right - You have to be in Eve prograde because it is slower then You on Kerbins escape, and You need to change relative velocity to Eve. It means fly Eve prograde (which is not possible) so You extract this proportionaly from Your speed against the Sun.

On 12/27/2020 at 1:03 PM, InfernoSD said:

That was until I tried timing a launch for Moho last week and failed repeatedly to get an intercept anywhere nearby.

Because it relative speed that would please You is more complicated because of it period to this speed.

On 12/27/2020 at 1:03 PM, InfernoSD said:

Moho defies this relationship by being 108 degrees "in front of" Kerbin despite being in a much lower orbit. I had to look this up online and reschedule my launch.

Yes - its relative speed close to the Sun draw a bit on this curve.

On 12/27/2020 at 1:03 PM, InfernoSD said:

Now that I know what the actual transfer windows look like, I'm left wondering. Is this principle I came up with just a coincidence resulting from how far apart these bodies are from Kerbin? Or was I nearing on some actual method?

From the Sun. From reference. You will get same result if trying to intercept for docking object in LKO by vessel behind Minmus.

On 12/27/2020 at 1:03 PM, InfernoSD said:

Or could there be any other way to eyeball transfer windows?

Yes. But there is no reason to use them. There are cheaper paths, but longer with gravity asist.

And If You would go hard way try inclined aproach against plane on which most planets are.

On 12/27/2020 at 1:03 PM, InfernoSD said:

And does the real solar system exhibit similar launch windows to Mars and Venus?

If we ignore KSP constants (due to memory and precision) - yes in all manner.

 

In KSP You can ignore transfer windows and go eliptical because dV to get away from Kerbin is extremaly low, and Yopu can refuel in space. We just do not have this oportunity in real world yet - gravity is not about to change, but refueling in orbit probably do.

 

 

 

 

[Superfluous J]

I've said this a few times in the past 8 years.

You should be able to make "planning nodes" that function exactly like maneuver nodes, but are on a planet's (or moon's) orbit. With them, it'd be as easy to find a Hohmann transfer as it is to do a rendezvous between 2 ships orbiting Kerbin.

As a bonus, the game could show you when and how to burn on your ship to most closely match that node.

Learn by doing, plus useful information. Win-win.

[king of nowhere]

i agree that this game gives too little information. i was very knowledgeable in astrophysics before, and yet i hit several walls upon playing and had to rely on this forum and spam questions for a few weeks to work out stuff that i knew was possible, but i just had no idea how to make in the game.

the one mitigating factor they have, though, is that space is so complex that there is no way to convey all the informations needed in a succint manner. and too much information can be even worse than not enough information. already the tutorials give you a ton of information, and there's no way you can remember it all at once. And have you looked at the control interface menu? there are so many options, i'm not even sure what half of those do!

i just don't see much chances to fix this game's steep learning curve.

no wait, i see one perfectly good way: whenever you have a doubt, go ask some expert. which is basically what the forum is for. but besides having the forum to ask, i see no other way to make the game easier to learn.

[OHara]

3 hours ago, HebaruSan said:

Imagine if you could go to map view and click a special new button to create a "meta maneuver node" on your parent body's orbit rather than on your own craft's orbit.

This does seem to be the most natural way to extend the existing KSP tools. 

In current KSP, we can cheat a craft, call it 'marker',  into a solar orbit very near Kerbin's (or the starting planet's) orbit, and create your "meta maneuver node" to reach the destination planet.   When the starting planet has reached that node, and real craft should be in orbit setting its ejection node, and it helps here to set the 'marker' craft as target, so we see both crafts' orbits after their maneuver nodes.  We match the real orbit to the 'marker's by eye, then target the destination planet to refine

A new player seemed to adopt this method naturally:

Edited January 4, 2021 by OHara

[HebaruSan]

5 minutes ago, OHara said:

"meta maneuver node"

@Superfluous J's choice of "planning node" is probably clearer and less intimidating.

[Spricigo]

14 hours ago, OHara said:

 

From the same thread:

 

If someone is more interested in the calculation I also found this old forum post: 

 

[Dr. Kerbal]

If you are saying that going interplanetary requires mods, it doesn't. I have been to every planet in KSP by just using the Internet and searching up the transfer windows.

[Dr. Kerbal]

Annd as a bonus I play on consoles so I play on older versions of KSP. I got to every planet and moon in 4 moths and I think i even got the ESA update and patch 13 or whatever.

[HebaruSan]

On 1/4/2021 at 12:21 PM, Superfluous J said:

You should be able to make "planning nodes" that function exactly like maneuver nodes, but are on a planet's (or moon's) orbit. With them, it'd be as easy to find a Hohmann transfer as it is to do a rendezvous between 2 ships orbiting Kerbin.

As a bonus, the game could show you when and how to burn on your ship to most closely match that node.

In case anyone is still interested in this thread, I finally got around to working on this concept. I know the OP is about stock, but this is how I picture stock could work without specialized transfer calculators if they decided to try to solve this problem: