Maneuver nodes innacurate

[[email protected] 1]

For some reason when I use the maneuver node I never end up where I’m supposed to be.

 

For example, when I used maneuver node it said my apoapsis is going to be 90km and my periapsis is going to be 85km, but I ended up with 242k apoapsis and 45k periapsis.

[Streetwind 4,947]

Posting a video is always a great way to get help

In this particular case, you are too close to the maneuver node by the time you start to throttle up. In other words, the maneuver node is too large for you to complete it at that moment in time. In yet other words, your engine does not have enough thrust. All of these are true.

But as CBase already pointed out, these are not the underlying cause. The actual cause is that your ascent is much too steep. You spend too much effort into going up, and not enough effort into going sideways. As a result, by the time your apoapsis reaches your desired altitude, you need to catch up to all this going sideways business. But there is so much to catch up to, you'd have to bring a really big engine and accellerate really hard to make it in time.

If you go sideways more during ascent, then the maneuver node you have to make to insert into orbit will be smaller. Additionally, it will also take you longer to push the apoapsis up to your desired altitude, which means you spend more time going sideways more, which yet again reduces the size of the final maneuver required. With practice, you will be able to get your final insertion maneuver down to less than 100 m/s.

So, how do you know how much sideways is enough sideways? Experience, mostly. But, here's a handy rule of the thumb: by the time your rocket has turned halfway over (45 degrees towards the horizon), you want to be somewhere between 10km and 12km in altitude. Don't sweat it if you don't hit it exactly; it is a rule of the thumb, not a law of physics. The exact "perfect" trajectory varies from rocket to rocket anyway, based on its thrust-weight ratio (TWR). The rule of the thumb assumes a launchpad TWR of roughly 1.4 to 1.5, which is considerd a good practices value for many kinds of rocket. If your launch stage has more oomph than that, turn over harder, and pass that 45 degree angle at 10km or even below; if your rocket barely manages to struggle off the ground, turn less aggressively, and go to 15 km or maybe even more for your 45 degree milestone.

 

Edited August 3, 2020 by Streetwind

[CBase 89]

Maneuver nodes show where you end, if you apply all forces instantaneous. This means you would need a incredible powerful engine to burn only fractions of a second. This is of course not possible. Normal burn durations are several seconds up to 1-2 minutes. If you need a longer burn for interplanetary travel you might want to split it into 2 or 3 nodes. As a rule of thumb you should never plan to burn longer than 1/16th of your orbital period.

And it is important to properly time your burn. If you start when the count down is at 0s, which means you reached the node position, in average all of your forces are behind the planned position. But you want them in average to be at the node. So if you have a burn duration of 60s as example, you want to start 30s before the node and stop 30s after the node. This way your average actual forces align with the planned node and you should end in approximately right orbit.

 

[Streetwind 4,947]

Welcome to the forums, @[email protected]

Judging by the fact that the periapsis drops, this looks very much like either a timing or an alignment issue. For example, making a node that's ten minutes into the future, and then immediately lighting the engines. Or, starting the burn at the correct time, but pointing in the wrong direction.

In either case, try looking a bit more closely at what the user interface is trying to tell you. This can be tricky if you're very new to the game, because there is a lot of information presented to you. And sometimes, there are things that us veteran players take for granted, but newcomers might not find immediately apparent. For example, did you know that when you make a maneuver node, an additional marker (colored dark blue) appears on the navball that tells you in which direction you must point? And that there's a timer at the bottom right side of the navball that tells you when the correct time to light your engines is? Perhaps you like flying with the navball collapsed, because you enjoy the view more that way. That would certainly lead to missing some of this information.

[[email protected] 1]

8 hours ago, CBase said:

Maneuver nodes show where you end, if you apply all forces instantaneous. This means you would need a incredible powerful engine to burn only fractions of a second. This is of course not possible. Normal burn durations are several seconds up to 1-2 minutes. If you need a longer burn for interplanetary travel you might want to split it into 2 or 3 nodes. As a rule of thumb you should never plan to burn longer than 1/16th of your orbital period.

And it is important to properly time your burn. If you start when the count down is at 0s, which means you reached the node position, in average all of your forces are behind the planned position. But you want them in average to be at the node. So if you have a burn duration of 60s as example, you want to start 30s before the node and stop 30s after the node. This way your average actual forces align with the planned node and you should end in approximately right orbit.

 

This is what I always do, it still doesn't work.

[[email protected] 1]

7 hours ago, Streetwind said:

Welcome to the forums, @[email protected]

Judging by the fact that the periapsis drops, this looks very much like either a timing or an alignment issue. For example, making a node that's ten minutes into the future, and then immediately lighting the engines. Or, starting the burn at the correct time, but pointing in the wrong direction.

In either case, try looking a bit more closely at what the user interface is trying to tell you. This can be tricky if you're very new to the game, because there is a lot of information presented to you. And sometimes, there are things that us veteran players take for granted, but newcomers might not find immediately apparent. For example, did you know that when you make a maneuver node, an additional marker (colored dark blue) appears on the navball that tells you in which direction you must point? And that there's a timer at the bottom right side of the navball that tells you when the correct time to light your engines is? Perhaps you like flying with the navball collapsed, because you enjoy the view more that way. That would certainly lead to missing some of this information.

Here's an example of what i tried to do:

https://streamable.com/ruxbhk

 

[CBase 89]

Thank you for that video !

Actually if you look close at the start of the burn (around 0:33s in video) you see that estimated burn duration is 2 minutes however only 53s are left until apoapsis. Yes this is pretty close, but hey this is rocket science ! And at 2:38 in the video you reached an 80x58 km (sub-)orbit, which is actually pretty good . Just after it you messed it a little up by trying to correct right now by following the manoveur target.

If you are around 10 m/s dV left, but already 1 minute behinde the node time, corrections on the original nodes (as displayed by the moving blue cross) are not the best choice.

Remember fixing periapsis is best done at next apoapsis. So accept the node as executed once it is that close and rather create an new node to correct your orbit. This might even mean that you coast quarter or half around Kerbin. But hey did you watch the Crew Demo return on this weekend ? They streamed 11 hours while they were constantly doing small corrections on multiple orbits. This is traveling in space. Luckily Kerbal space has some time warp to skip long coasts

Oh and a minor hint for your next ascent: 1400 m/s circulization burn indicates you did ascend mostly straight up. If you turn little more sideways early in ascent you burn more horizontal and need less time to circulize. Which makes it easier to time. But really only do very small corrections, you will see that drag and changed ascent curve change big even for small turns early.

[Laie 1,408]

2 hours ago, [email protected] said:

Here's an example of what i tried to do:

https://streamable.com/ruxbhk

Maneuvers do not work very well if you go around a bend while doing them.

OK, being in orbit means you're always going round the bend... but it's still a matter of degree, quite literally. Over the time it will take you to perform the burn, you original trajectory changes from going steep up to steep down, describing perhaps a 120° change of direction. That's a lot.

CBase already mentioned the problem in his very first reply:

11 hours ago, CBase said:

As a rule of thumb you should never plan to burn longer than 1/16th of your orbital period.

IMO, the word "period" should be dropped from that rule. The problem isn't how long it takes, but the underlying change of your prograde vector. In your example, the burn extends over about 1/3rd of an orbit.

[CBase 89]

3 hours ago, Laie said:

The problem isn't how long it takes, but the underlying change of your prograde vector. In your example, the burn extends over about 1/3rd of an orbit.

Actually the rule is about center of gravity vector change and less about prograde which would land you in different target orbit than planned. So yes the rule of thumb only applies to mostly circular orbits.

At highly eccentric orbits - like the one in video after a very straight ascent - the craft spends considerable time near the apoapsis. Therefore gravity direction hardly changes and you can plan longer burns.

I did this kind of LKO ascents with extreme non aerodynamic payloads, you can do it in a single symmetric burn. You do loose of course some efficiency when node marker and prograde are not aligned, but as a starter that is perfectly fine. His problem is really about starting 7s too late and following the node marker to fine tune orbit at wrong point in orbit.

[Spricigo 1,146]

13 hours ago, CBase said:

..reached an 80x58 km (sub-)orbit, which is actually pretty good . Just after it you messed it a little up by trying to correct right now by following the manoveur target.

Actually the98x68k orbit at end of execution is pretty close to the predicted 94x72, specialy if we consider the small inacuracies in piloting. Mind you, that is easier to notice when not in a hurry to set up and execute a maneuver.

 

15 hours ago, Streetwind said:

Don't sweat it if you don't hit it exactly; it is a rule of the thumb, not a law of physics. The exact "perfect" trajectory varies from rocket to rocket anyway, based on its thrust-weight ratio (TWR). The rule of the thumb assumes a launchpad TWR of roughly 1.4 to 1.5, which is considerd a good practices value for many kinds of rocket. If your launch stage has more oomph than that, turn over harder, and pass that 45 degree angle at 10km or even below;

I still think that rule of thumbs needs to account for Moar Thumpers.  

For reference: Bester, launchpad TWR2.5, 45º at ~800m, 25º at 10km