Burn Time & Maneuver Nodes

[rodion]

Not sure if this has been covered elsewhere but I've read enough threads without seeing anything so I'll ask. When one sets a maneuver node and then performs the prescribed burn at t-0s, especially in short-period places like low planetary orbit and with long burns, the result is often a distortion in the shape and position the trajectory compared to the one set. If I can estimate though, that a burn will take e.g. 90 seconds, if I start burning towards the target node at t-45 and finish at t+45, will that minimize the distortion in orbit? I've tried it a few times but I could never exactly tell if I was making the issue better or not. Is it maybe more subtle and less straightforward than that?

[el_coyoto]

In my experience, yes this is the way to do it : if you have a 90 seconds burn, you start at T-45 s and end at T+45 s (if all goes well).

The problem with long burn times is that you have to keep the ship precisely pointed for a longer time, thus sometimes resulting in an accumulation of guidance errors (I guess).

[Somerled]

Yup, you got it. Half the burn time before, half after is a solid rule of thumb.

Alternatively, you can start at t-90s plus a bit in your example and adjust for distortion during the burn so that everything lines up much better at t-0s, not t+45s.

For example, you're approaching periapsis and attempting to circularize. If you burn directly at your target vector you'll lower your periapsis and distort your orbit from desired (a little or a lot, depending on the maneuver and craft). But if you pitch up just a little bit, you can maintain your periapsis altitude at the cost of taking a little bit longer to slow down. If your periapsis rises too far, pitch down. It's less fuel efficient than the half-&-half way, but it makes up for the fact that you planned your maneuver assuming an impulse burn, which is unrealistic.

[rodion]

The problem with long burn times is that you have to keep the ship precisely pointed for a longer time, thus sometimes resulting in an accumulation of guidance errors (I guess).

Yeah, even worse when I'm moving complex payloads and the awkward center of mass defeats the power of any ASAS to keep the ship on course.

[el_coyoto]

Yeah, even worse when I'm moving complex payloads and the awkward center of mass defeats the power of any ASAS to keep the ship on course.

Very annoying, indeed...

What I sometimes do is adding a Poodle or a more powerful engine to my stack, and use it in conjunction with the weaker/more efficient engines during the ejection burn. I then jettison the "kick" engine once it is not needed anymore (ship is light enough, precision is not so important anymore for course corrections, etc).

It's a trade-off : it hurts the delta v of the ship a bit, but it makes it easier to fly. Maybe it could also use the added gimbaling and power to better balance your ship (depending on how unbalanced it is...)

EDIT : I'm so proud I recognized your avatar! Sergei FTW!

Edited July 29, 2013 by el_coyoto

[Amoun]

I had the very same question.

I have been paying attention to the length of the burn.

If the burn is as you say, I would start 15 seconds before (no math), short burns (60 seconds and less) I start at t-0 and it works out.

For longer burns I try to split it like you do (Ie: 90 second burn: t-45, t+45)

I also noticed that if I play around with timing in that way, the direction of your burn (the blue cross-hairs) would get distorted as you are catching up with m/s. If I follow it, my orbit becomes almost exact as the one set.

I have not done too many experiments to give you precise findings but I hope this helps even a little.

[rodion]

I'm very happy with these answers and now I'm not sure if I can change the thread tag to "Answered" manually. Does a mod or someone actually need to do that? I don't know...

What I sometimes do is adding a Poodle or a more powerful engine to my stack, and use it in conjunction with the weaker/more efficient engines during the ejection burn. I then jettison the "kick" engine once it is not needed anymore (ship is light enough, precision is not so important anymore for course corrections, etc).

I also considered this very same approach, although I've never been confident enough with my ÃŽâ€v margins to ever try it. Perhaps now I should, since I can launch missions with more fuel than I actually need at this point.

EDIT : I'm so proud I recognized your avatar! Sergei FTW!

That rules, at 64 pixels avatars become kind of stupidly small and vague. Even now his strong, shining features defeat all obstacles.

[metaphor]

Burning your maneuver node half before and half after the time is a good rule of thumb.

But if you want to be more precise, you have to take your burn delta-v into account. Since you're accelerating during the burn, you're arriving at the node faster than if you weren't burning. So to counter that, if it's a large delta-v maneuver compared to your orbital speed, do more than half the burn before node time and less than half after. For example if you have a 50 second 1000 m/s prograde burn and you're moving at an orbital speed of 2000 m/s, I would start the burn about 30 seconds before the node time instead of 25 seconds. That way you will arrive at the same place where the node is exactly halfway through your burn. If you're slowing down on your burn (retrograde), do less than half the burn before node time and more than half after.

[xylotism]

Burning your maneuver node half before and half after the time is a good rule of thumb.

But if you want to be more precise, you have to take your burn delta-v into account. Since you're accelerating during the burn, you're arriving at the node faster than if you weren't burning. So to counter that, if it's a large delta-v maneuver compared to your orbital speed, do more than half the burn before node time and less than half after. For example if you have a 50 second 1000 m/s prograde burn and you're moving at an orbital speed of 2000 m/s, I would start the burn about 30 seconds before the node time instead of 25 seconds. That way you will arrive at the same place where the node is exactly halfway through your burn. If you're slowing down on your burn (retrograde), do less than half the burn before node time and more than half after.

This is pretty good advice. I try to start my burns early to make sure I don't end up finishing far too late and mess up the whole maneuver.

I have a few tips though:

* The maneuver reticle on your navbar adjusts on the fly. Try to match it perfectly regardless of your timing, as long as you stay on that target and deplete the delta V bar to 0m/s, you should end up at the right orbit.

* For longer burns it's often more efficient to do some of the maneuver on your first orbit and do the rest on the second pass, because the farther out you start/end your burn the less fuel efficient it will be.

* Turn off your SAS as the burn comes to an end, so it's easier to stay on target as the blue reticle moves.

* Press X to kill throttle immediately. If you're throttled all the way up it can be hard to throttle down manually and keep the maneuver intact.

[Frederf]

I've found that the burn isn't perfectly symmetrical around the ideal "instant impulse" point if one is approaching a periapsis since acceleration before the point tacks on m/s in addition to the "falling" acceleration which has a multiplicative effect on total velocity change. I've found roughly 33% before and 66% after results in an exit that is closer to that of an instant burn.

[KerbMav]

What about the incorrect prediction of burn time?

[rryy]

What about the incorrect prediction of burn time?

The game calculates burn time based on the last time you accelerated. So if your last burn was with your previous stage or at a low throttle, the game will calculate the burn time based off of that, not off of your current stage. Once you begin the burn, the game will recalculate the burn time based on your current acceleration.

[KerbMav]

The game calculates burn time based on the last time you accelerated ... if your last burn was with your previous stage or at a low throttle ...

Hah! Thank you!

[ASnogarD]

Another trick I use ... and it seems to work is to start your burn at t = 0 minus est. time but half the thrust ... for example, you have a 90 sec burn , so 90 sec before the node burn at HALF thrust (obviously this is for small burn times, those under 5 min), the result is the same as if you went full throttle at the half est. time but you have more control and you also incidentally get some scope to cope with bad estimated times (basically if it said you had a 60 sec burn but change to 1 min... you can adapt by going full throttle instead of half).

[el_coyoto]

I'm very happy with these answers and now I'm not sure if I can change the thread tag to "Answered" manually. Does a mod or someone actually need to do that? I don't know...

I think you need to manually add '[Answered]' to your post title.

Cheers!