Breaking up long burns into multiple burns - How?

[ibanix]

So I've got a ship ready to go to Eve, using nukes. The delta-V needed is only about 1000, but the burn time with the nukes is just over 4 minutes. My ship is in a 100km parking orbit, and a 4-minute burn is going to mean some extra dV spent in order to compensate for not being entirely prograde during the burn.

I've seen that other people break long burns into multiple small burns with a complete orbit in between. How do I figure out the timing of these burns?

Thanks in advance.

[4x4cheesecake]

Maybe it's just me, but I wouldn't split a 4 min burn.Thats just not long enough.

But if you really want to split that burn, don't bother about timing that much, because the timing is getting less relevant when setting up a long distance trajectory, it's more about precision regarding the burn direction and duration.

So what I'm doing when splitting a burn is: setting up one maneuver that brings up an encounter and then perform the burn but pretending, it is only a 4 min burn (what a coincidence^^), so 2 min before and 2 min after the node, pointing to the maneuver marker on the naveball. I'm actually not sure if it is possible to add just +1 orbit to the node, I'm always getting wrong deltaV numbers when I've tried but I'm using the maneuver node evolved mod, maybe thats the problem.  Due this problem, I just delete the node a place a new one direct at Pe (that is exactly where the old node was), and setting up the maneuver like before. Repeat until I have reached my desired orbit/encounter. Maybe not a perfect strategy but its working fine for me

If you don't belive me when saying the timing is not a big deal, you can actually try it: While in LKO, set up an maneuver to encounter eve and write down the required deltaV. 100km LKO orbital period is ~30min, so add 2 more orbits to your node and change the vectors to get an encounter again. The difference between your new deltaV and your old deltaV should be some where around 0.5 or 1m/s. So you need 1m/s more deltaV if you are 1 hour late on your transferwindow.

If you want, you can split your 4min burn into 4x 1min burn and starting the first burn at the perfect transferwindow, be ~8h late on your transferwindow when finishing the final burn which will cost you a tiny amount of deltaV to correct that, but you will safe a lot more fuel by using the oberth effect more efficent. I guess, you can even split the time for the transfer window like a maneuver node time (start 4h early and finish 4 hours late) but this will only save some fuel in your correction, which is already just a tiny amount.

 

 

[Goody1981]

Not sure if there's a mathematical way, but my approach is to kind of eyeball it (possibly with some quick saving in between).

For example, if you're talking about a 4 min burn, maybe try splitting it up into 2 x 2 min burns. The first burn might put your orbit out just past Minmus (obvs depends on your ship), then you just time accelerate, come back to your initial burn point (which is now the periapsis of your orbit) and do the second burn. Of course, if you start your first burn at exactly the time you want to depart for your perfect Eve transfer window, then orbiting out to Minmus and back before completing the burn is going to make you a few days late.

So you can either:

- not care, a few days either side of a transfer window won't make too much difference

- do the first burn a few days earlier (this is where quick saving can help to figure out the timings) so that your second burn will be on the correct date of departure

 

You can also mess around with different numbers of burns, e.g. splitting it into 3 x 1 min 20 s burns will allow each burn to be shorter, but because you'll be doing two extra orbits before your final ejection, you'll have to start the first one even earlier. The burns also don't have to be evenly split. If you've decided 4 min is too long to do in one burn, but maybe you're OK with 3 min, well then do a 1 min burn (which might only increase orbital period by a few hours rather than a few days, so you don't even really have to worry about starting earlier) and then make your second burn the remaining 3 mins for ejection to Eve.

BUT what you MUST make sure of, no matter how many burns you do, is that you DON'T leave Kerbin's sphere of influence until the final burn. In my first example of the 2 x 2 min burns, if your first burn put you into solar orbit (rather than stopping just past Minmus) then you won't be coming back to your burn point, and therefore you can't finish the burn.

[ibanix]

Thanks for replies. I know 4m isn't a terribly long burn, but I'd like to add this "technique" to my toolbox for when I try to send a 100t station to Jool and need to split a 90m long burn, for example.

I'd love to hear from anyone who has done split up a super-long burn into many of them, and how they timed them so as to still get the transfer window correct.

[4x4cheesecake]

There are some limitations for extrem long burn times, first of all:

3 hours ago, Goody1981 said:

you MUST make sure of, no matter how many burns you do, is that you DON'T leave Kerbin's sphere of influence until the final burn.

If your final burn is still very long, you will be limited by the altitude of your Pe. The final burn will lower your Pe and you totally don't want to go trough Kerbins atmosphere (or through Kerbin itself). In this case, you have to raise your start orbit but with an higher orbit, you will loose profits from the oberth effect.

I guess, your "100t payload to jool with 90m burntime" example is intended to be exaggerated but lets take a look at the numbers: Starting in 100km LKO, you will need ~915m/s deltaV to raise your Ap to the edge of Kerbins SOI. Thats not even 1/2 of the required deltaV to perform a Hohmann transfer to Jool (will need ~2000m/s deltaV, according to Alex's Moon launch window planner.) So you will still have ~45min burn time for your final burn and I actually doubt that it is possible to do a 45min burn with a Pe at 100km. In my optinion, a final burn shouldn't be longer than 5min (but thats just my opinion and gut instinct, maybe someone did the math and can provide an more founded answer to this question). You can still raise your Pe and start your journey from a higher orbit but your final burn will be so inefficent that you will end up with more deltaV, despite all the savings by the oberth effect.

So you can raise your TWR or go for an slingshot around Eve, but a slingshot is even harder to time than splitting a long burn (to reach Jool you would need multiple slingshots bwtween Eve and Kerbin). Recently I read a pretty good tutorial to do (multiple!) slingshots and how to find the timings, if I can find the link, I'm going to post it here for you.

If you still want to go with low TWR and a long final burn, I would recommend to rais your Ap up to Minmus (~900m/s deltaV, orbital periode ~17 days) or above to the edge of Kerbins SOI (~915m/s deltaV, orbital periode ~40 days) and start your burn several days before the transfer window opens. How many days? lets see: I would highly recommend a mod like Kerbal Engineer or MechJeb to get some fancy readouts (like post-burn orbital period for planned maneuvers), so you can easily add the orbital periode for each single burn. The sum of all orbital periods to raise your orbit up to Minmus or the edge of the SOI is the time you want to start before your transfer window. Then, place another maneuver node but add as many orbits as possible, so this node will be as close as possible to your transfer window, play around until you get an encounter with your target body: Where ever the node is, that is the position you want to start from. Now you got the position and the time for your first maneuver. But still keep in mind, the timing is pretty irrelevant if your target is far away, so dont worry too much about that

[bewing]

If you burn prograde, your Pe doesn't drop.

 

[4x4cheesecake]

43 minutes ago, bewing said:

If you burn prograde, your Pe doesn't drop.

 

Correct but you will not reach the exact orbit planned with the maneuver node. The common way to perform a maneuver is doing half the burn before and the other half after the node (you know that). But your Ap will not raise constantly over time and your orbit will be shifted radial (I'm pretty sure, you know that too).

[Rauko]

Does really starting your burn 2 minutes before the node induce a cost in dv? I mean, you are adding a vector to your movement, as long a you do it "simmetrically", AND you are pointing the maneuver direction the whole burn, it just happend that "the other engine" that moves your ship (that is, kerbin) is adding its own vector, but half time in one direction, half time in the other)

The biggest part of your savings will come from oberth.

Am I wrong?

How big is the oberth effect? With the Jool burn example, with a burn of 2000dv, how much would we save doing it in two burns? Is it around a 1%? A 20%?

[ibanix]

30 minutes ago, Rauko said:

Does really starting your burn 2 minutes before the node induce a cost in dv? I mean, you are adding a vector to your movement, as long a you do it "simmetrically", AND you are pointing the maneuver direction the whole burn, it just happend that "the other engine" that moves your ship (that is, kerbin) is adding its own vector, but half time in one direction, half time in the other)

How big is the oberth effect? With the Jool burn example, with a burn of 2000dv, how much would we save doing it in two burns? Is it around a 1%? A 20%?

Scott Manley had a youtube video on this, and I believe he got something like a 20% difference in dV when everything was factored in. 

The main issue with long burns in low orbits (ie, a 100km orbit) is that you are burning off-axis to where you need to burn. The optimal burn would be an instantaneous burn (zero second) that gave you 100% of the dV you need for the orbit change. Every second you're not doing that, you're spending extra dV in order for the final orbit to be what you would get, if it was a zero-second burn.

So it turns out there is a plugin that handles splitting up burns - Maneuver Node Splitter - but it has not been updated beyond KSP 1.2.2. 

2 hours ago, 4x4cheesecake said:

If you still want to go with low TWR and a long final burn, I would recommend to rais your Ap up to Minmus (~900m/s deltaV, orbital periode ~17 days) or above to the edge of Kerbins SOI (~915m/s deltaV, orbital periode ~40 days) and start your burn several days before the transfer window opens.

Sure, that's a possibility. But it wastes more dV to get up into that orbit and then burn out. I know it's possible to use multiple burns, I just can't figure out the math involved. Hasn't *someone* done this?

Edited May 9, 2018 by ibanix

[Goody1981]

1 hour ago, ibanix said:

Sure, that's a possibility. But it wastes more dV to get up into that orbit and then burn out. I know it's possible to use multiple burns, I just can't figure out the math involved. Hasn't *someone* done this?

Pretty sure he's talking about splitting the burn, not burning up that far and then...burning again at Ap? It's not wasting dV.

[RizzoTheRat]

2 hours ago, ibanix said:

Sure, that's a possibility. But it wastes more dV to get up into that orbit and then burn out. I know it's possible to use multiple burns, I just can't figure out the math involved. Hasn't *someone* done this?

There are 2 components to an interplanetary burn.  First you need the dV to escape from Kerbin, which puts you in to the same orbit around Kerbol as Kerbin is in.  Then you need the dV to do the Hohmann transfer from Kerbin to Jool.

@OhioBobdid a great explanation of this in reply to my question about a Mun return, but going from Kerbin to Kerbol orbit is the same maths as going from Mun to Kerbin orbit.

When you split the burn you're getting a chunk of the escape velocity burn in early, but obviously can't add more than that without leaving Kerbin orbit.  What I like to do with big ships is use a tanker/tug to help haul them out to near the SOI limit, refuel, and then separate from the tanker for the interplanetary burn on the next orbit.

If you work out the orbital period of an eccentric orbit from KLO to near the SOI, you know you need to do the initial burn that long before the ideal window for the transfer, so know how many orbits early to start the first burn, eg if your LKO orbit has a period of 45 minutes and your eccentric orbit is 6 hours*, you need to do the initial burn 8 orbits early. 

 

 

* completely made up numbers, it'll be way more than that if you're getting towards the SOI

Edited May 9, 2018 by RizzoTheRat

[Rauko]

The more I think about it, the less I understand it.

I can see the Oberth effect the more speed you already have, the more speed you gain per kg of fuel spent.

 

But does is mean that it is beneficial to raise apoapsis and then burn again in your next periapsis? I dont see it.

Let me explain.

 

You are in lko flying at 2200m/s. You need a burn of 2000m/s to reach your destination, and KER says you have exactly 2000dv remaining.

 

You think: wait! I will get advantage of oberth effect, I will raise my apoapsis to raise my speed at periapsis. I spend 1000 dv to get a eliptical orbit, and in my next periapsis I will burn. In the periapsis, the ship is now faster, say 3200m/s, and due to oberth, I will need less fuel to reach my desired destination, leaving me with, say, 200dv remaining in my KER readout.

 

I gained those 200dv because I was faster when I did the second burn. BUT when I finished my first burn I WAS ALREADY THAT FAST! (due to that velocity, my orbit was eliptical at that very moment), so I would have take advantage of Oberth just by not stoping my first burn. (IE, KER readout would have anyway ended showing 200dv remaining)

 

Where did I lost contact with reality?

[RizzoTheRat]

Because the burn takes a finite duration, rather than the instant change in velocity that a manoeuvre node simulates, a lot more of the burn takes place at a higher altitude, and therefore lower speed.  Splitting the burn over 2 orbits means you're burning at a lower average altitude, and therefore higher average speed.

I think. 

Oberth is quite confusing.

[Rauko]

11 minutes ago, RizzoTheRat said:

Because the burn takes a finite duration, rather than the instant change in velocity that a manoeuvre node simulates, a lot more of the burn takes place at a higher altitude, and therefore lower speed.  Splitting the burn over 2 orbits means you're burning at a lower average altitude, and therefore higher average speed.

I think. 

Oberth is quite confusing.

Mmm... BUT, if you are in lko and burn once, the first half of your burn is made at your minimun altitude.

And if you do a double burn (half time each one) you are at minimun altitude only in the first half of first burn. Average altitude is bigger  (maybe average speed is also bigger, and thats the trick)

[RizzoTheRat]

You should be able to keep the periapsis at about the same altitude for both burns.  So for a 4 minute single burn your fastest speed, ie the end of the burn, happens 2 minutes after periapsis,  whereas with a split burn it's only 1 minute after periapsis so should be at lower altitude 

[Kryxal]

I like to do a burn of about 430 m/s if I'm splitting a node ... more precisely, a burn to get an orbital period of 90 minutes.  It lets you set things up better, because you can just execute the first burn 3 orbits before you would have done the "full" burn.

[OhioBob]

15 hours ago, ibanix said:

So I've got a ship ready to go to Eve, using nukes. The delta-V needed is only about 1000, but the burn time with the nukes is just over 4 minutes. My ship is in a 100km parking orbit, and a 4-minute burn is going to mean some extra dV spent in order to compensate for not being entirely prograde during the burn.

I've seen that other people break long burns into multiple small burns with a complete orbit in between. How do I figure out the timing of these burns?

Thanks in advance.

What I like to do in that case is perform two burns exact one day apart.  The first burn is performed at the location and time of day that you want to perform the ejection burn, but one day earlier.  You give the vessel enough delta-v to put it in an orbit with a period of 6-hours.  This means that one day later it returns to the correct spot at the correct time to finish off the ejection burn.  From a parking orbit of 100 km, the first burn should require a delta-v of about 763 m/s.

BTW, a 4 minute burn isn't really all that bad.  You could probably do that in a single burn without too much loss of efficiency.

[bewing]

2 hours ago, Rauko said:

I can see the Oberth effect the more speed you already have, the more speed you gain per kg of fuel spent.

I think a better way of looking at it is more like:

When your orbit is not circular, you lose speed to gravity losses as you climb. The longer it takes you to climb, the more speed you lose. So the Oberth Effect is all about getting you out of the gravity well ASAP. This allows you to carry as much speed as possible along with you. So if you start with 700 m/s and add 2000 m/s and it takes you 20 days to leave the SOI, you will be going very slow once you are out. If you start with 2400 m/s and add 2000 m/s and it takes you 20 minutes to leave the SOI, then you won't lose much.

 

[OhioBob]

4 hours ago, RizzoTheRat said:

There are 2 components to an interplanetary burn.  First you need the dV to escape from Kerbin, which puts you in to the same orbit around Kerbol as Kerbin is in.  Then you need the dV to do the Hohmann transfer from Kerbin to Jool.

@OhioBobdid a great explanation of this in reply to my question about a Mun return, but going from Kerbin to Kerbol orbit is the same maths as going from Mun to Kerbin orbit.

 

It sounds like you misunderstood me.  I do not recommend two separate burns in the way that you describe.  It is most efficient to perform a single burn when near the planet.

While I separately calculated the delta-v needed to change the periapsis after escaping, this was done only to determine how much hyperbolic excess velocity is needed.  The burn that provides the hyperbolic excess velocity is performed near the planet in a single burn.  You burn enough to increasing the vessel's velocity to escape velocity, and then you continue to burn to give it some extra energy.  This extra energy is left over after you escape in the form of hyperbolic excess velocity.

[Rauko]

I'll try and gather some empirical data tonight.

[Corona688]

9 hours ago, 4x4cheesecake said:

Correct but you will not reach the exact orbit planned with the maneuver node.

I've done my most accurate long burns that way - accurate in terms of reaching the target if not perfectly following the dotted line.  It's also more efficient.  Burning prograde mucks up a burn's time-symmetry though, since you'll be pushing your peristasis upwards for the entire burn.  I split time 30/60 across them instead of 50/50.

Edited May 9, 2018 by Corona688

[4x4cheesecake]

13 minutes ago, Corona688 said:

I've done my most accurate long burns that way - accurate in terms of reaching the target if not perfectly following the dotted line.  It's also more efficient.  Burning prograde mucks up a burn's time-symmetry though, since you'll be pushing your peristasis upwards for the entire burn.  I split time 30/60 across them instead of 50/50.

You may wat to share your strategie since the original topic is about splitting long burns and being accurate / getting the right timing. I'm quiet interessted aswell (splitting the time 30/60 cannot be the whole trick, can it?)

[Corona688]

There is a little more to it but not much.

I watch the target indicator to tell if I've started too soon.  The target indicator always gets pushed away from the center of the navball by thrust.  But the travel of your orbit should overcome this and keep bringing it towards - and eventually past - the prograde indicator.  If the target indicator flies straight away into lala land, you started your burn too early.

Oh, and I burn during a 30-45 degree orbital segment around the node, more or less.  This is very much eyeballed, not a math thing.

Edited May 9, 2018 by Corona688

[Rauko]

16 minutes ago, Corona688 said:

There is a little more to it but not much.

I watch the target indicator to tell if I've started too soon.  The target indicator always gets pushed away from the center of the navball by thrust.  But the travel of your orbit should overcome this and keep bringing it towards - and eventually past - the prograde indicator.  If the target indicator flies straight away into lala land, you started your burn too early.

Sorry, when you say split 30/60 you mean before/after the node, yes? Not splitting in two separate burns.

[Corona688]

Oh, yes.  That's time splitting for a single burn, before the node and after a node.