[ANSWERED] How to Execute Extremely Long Burns

[March Unto Torment]

Hey, everyone!

So, I'm in the midst of planning a Jool-5 mission, for a craft which I'm currently assembling in KEO (it was originally a space station, hence the odd location for assembly). However, for... reasons, my total burn time on the transfer stage - that is, just the stage I'm using to get there from KEO, which has almost exactly enough dV to get to Jool from KEO (4,131m/s) - is sort of ridiculous.

That is, it's one hour, fifty minutes and twenty-nine seconds. Due to the nature of the mission (it'll all become apparent when I actually do the report), increasing TWR or decreasing craft mass are not available options.

Now, as I see it, my only real advantage is that I'm starting from KEO, which means that such a burn could (if done all at once) be completed in only 1/3 of an orbit, as opposed to taking ten thousand or so orbits if I was in LKO.

Any suggestions for where to go from here? Is splitting it up into multiple burns a viable option? If so, how many burns should I use?

Many thanks to anyone who can help out!

Best regards to all,

 - March Unto Torment

EDIT: Question has been answered! Thanks to the long-suffering efforts of Plusck, I now have a direct answer to my question (along with an absolutely gargantuan amount of auxiliary information on how to get to Jool). For any future Kerbonauts wanting to know specifically how to execute long burns without wanting to wade through my ramblings, bear witness to Plusck's genius:

Quote

 

For the burn from KEO outwards, I would highly recommend plotting it using the method I described earlier and which I used for comparison purposes, so I know that it works and is easy and accurate, and guarantees the best efficiency for the burn:

• plot a single burn to start with to get a Jool intercept. It should be about 2100 m/s. Note how long it will take (you may have to fire your engines for a second to get the numbers to show);
• drop a new maneuvre node that is more-or-less exactly this much time after the ejection burn, plus a couple of minutes of margin to be safe. Give it a certain prograde amount just to register. If all goes well, this will end up being the last of the series of burns you need to make to get to Jool;
• delete the first node, and correct the other one to give about 10 minutes of burn;
• now create a new node exactly 10 minutes before that final burn, give it 10 minutes' worth of burn;
• repeat, going 10 minutes earlier each time (and if ever it isn't quite 10 minutes earlier, make the burn for slightly less than 10 minutes...) until you get a Jool capture
• with 0.55m/s² acceleraton, you should have 6 burns plotted;
• check your ejection angle: since you gave a few minutes' margin at the end, it should be almost exactly the angle of the original single burn. If it isn't:

1. if the angle is too radially in with respect to the sun, reduce the amount of the first burn and increase the amount of the final one;
2. if too radially out, add another burn 10 minutes earlier and reduce or delete the last one

When you actually do the burns, each one should be exactly prograde (or close enough to make no difference), which is the most efficient. By plotting a series of prograde burns, you eliminate the error that creeps in to any long burn - both in ejection angle and in the actual fuel cost of the burn because of the off-prograde angle.

In fact, if your nodes have been accurately spaced 10 minutes apart with exactly 10 minutes of burn each time (and assuming that your TWR will not change much from start to finish) you could even forget about the nodes altogether: block off fuel tanks so that you have exactly the right total amount of fuel available for 2100 m/s (or whatever it is), have SAS align to prograde, start burning a couple of minutes before the first burn then go watch a movie. ; )

 

Edited April 9, 2016 by March Unto Torment

[Superfluous J]

With that long a burn, your only option is to do it really inefficiently and fix it after the main burn's done, preferably outside of Kerbin's SOI. You're not going to get any benefit from Oberth and in fact you're not going to even be able to do the burn near Kerbin, you'll do a full orbit (or two!) during the burn time. You will HAVE to split the burn, as you can't burn backwards in your orbit

The only way I see to do it is to do multiple burns at Pe to raise your Ap up far enough that you'll be up in that area for a couple hours, and then do the actual transfer burn up there. IOW, don't do the transfer burn in LKO, but instead somewhere up near Keostationary or higher.

Edited April 5, 2016 by 5thHorseman

[March Unto Torment]

12 minutes ago, 5thHorseman said:

The only way I see to do it is to do multiple burns at Pe to raise your Ap up far enough that you'll be up in that area for a couple hours, and then do the actual transfer burn up there. IOW, don't do the transfer burn in LKO, but instead somewhere up near Keostationary or higher.

Not sure if you missed it in my OP, but the main assembly's already been constructed in Keostationary Orbit. So from the sounds of what you're saying, that's one of the larger problems fixed. Question is, since I'm in KEO, can I do one single burn, or am I still going to need to do multiple? (Most dV maps are telling me that I need 3,800m/s or so to transfer to Jool from KEO, so I don't have much dV to spare when it comes to fixing up my orbit.)

[Superfluous J]

4 minutes ago, March Unto Torment said:

Not sure if you missed it in my OP, but the main assembly's already been constructed in Keostationary Orbit. So from the sounds of what you're saying, that's one of the larger problems fixed. Question is, since I'm in KEO, can I do one single burn, or am I still going to need to do multiple? (Most dV maps are telling me that I need 3,800m/s or so to transfer to Jool from KEO, so I don't have much dV to spare when it comes to fixing up my orbit.)

Oh I thought KEO was "Kerbal Equatorial Orbit" I suck at acronym.

KEO (hey I'm using it now!) is 6 hours, and you need to burn for 2 of those hours. That's about 1/3 of the way around. I'd say that's too much.

I'd personally be more comfortable burning at the other side of the orbit, raising my Ap up maybe double, and then doing the burn from there.

And again, that's super inefficient, and will cause your burn to be even longer.

No matter what you do, WATCH YOUR RESULTANT SUN ORBIT instead of the maneuver node as you near the end. Your maneuver node marker will be TOTALLY WRONG as you get farther and farther from it, and you're far better off just holding the line and watching your Sun Apoapsis.

Incidentally, splitting the burn and doing it in parts, each time at Pe raising Ap, isn't an option here as you'll need like 800m/s to leave Kerbin's SOI, and another 1000 or so to get to Jool. That 1000m/s would need to all get done at once, and would cost you about an hour of burn time. It's not QUITE as impossible from Keostationary, but it's still a pretty daunting burn. And out there you're not saving nearly as much from Oberth so not doing that isn't quite as bad.

Edited April 5, 2016 by 5thHorseman

[Plusck]

On the plus side, if you're starting from KEO you're actually not losing all that much. Sure, you're losing out on the Oberth effect, but at the same time the actual cost of the ejection burn can be lower at higher orbits. See here:

Those calculations do not mean that you're better to raise your orbit before the transfer burn (since that'll cost more overall), but it does show that if you're already at the given orbit, your burn will be optimal.

Unfortunately for you, the optimal orbit for Jool is apparently 371km... so quite a bit lower than KEO.

 

This means that looking at gravity assists might be a good idea - specifically from the Mun, maybe even from Duna.

 

Doing your burn over a third of an orbit is impossible - by the time you're halfway through the burn, your orbit will have changed radically and you'll already be close to a hyperbolic trajectory, and at some considerable altitude, so you'll never be able to get back on track to get the right ejection angle. This makes planning difficult, and you'll end up doing a significant amount of your burn in sun orbit anyway.

If I were in your position, I'd plot a number of burns, each for a maximum of ten minutes or so. KSP doesn't make this easy though...

The easiest is to plot burns progressively nearer to your craft (if KEP lets you...), so the first would be just behind your craft (maybe a fake one for much more than 10 minutes, to avoid confusion between pre- and post-burn orbits), then you'd drop another node 12 minutes or so before that one, then another before that, and so on. This method lets you get the numbers right first time, avoids jitteriness in the burn you're plotting (but not for the later ones) and makes it possible to get a reasonably accurate trajectory for a Mun flyby if you want (which, given your speed and angle, will probably only save you a couple of hundred m/s, the Mun just isn't big enough to help much on Jool-velocity escape burns). As you work back in time, you should begin to see where you need to start to get the right escape angle.

Well, that's what I would do in your position, but I've only very rarely had to deal with long burns on low-TWR craft.

[Superfluous J]

One other option is to do it the way that's wrong but that most new people do: Burn just enough to get out of Kerbin's SOI (I suggest using a Mun slingshot for this, it'll save you some of the dV you're wasting doing it this way) and then once in Sun orbit plot and execute the Jool transfer. An hour+ burn in Sun orbit isn't a problem as the fraction of the total orbit is so small.

If you go this route, just make sure you leave Kerbin's SOI somewhere on the front half, preferably heading a bit away from the Sun. So, heading in the direction you'll be going when you burn to Jool. That way, you don't re-encounter Kerbin during or after your transfer burn.

Edited April 5, 2016 by 5thHorseman

[March Unto Torment]

2 minutes ago, 5thHorseman said:

One other option is to do it the way that's wrong but that most new people do: Burn just enough to get out of Kerbin's SOI (I suggest using a Mun slingshot for this, it'll save you some of the dV you're wasting doing it this way) and then once in Sun orbit plot and execute the Jool transfer. An hour+ burn in Sun orbit isn't a problem as the fraction of the total orbit is so small.

That seems like the most viable way to do it without screwing up, although it does beg the question of how much dV I waste this way (I've only got a couple of hundred m/s to spare, if you'll recall).

[Superfluous J]

3 minutes ago, March Unto Torment said:

That seems like the most viable way to do it without screwing up, although it does beg the question of how much dV I waste this way (I've only got a couple of hundred m/s to spare, if you'll recall).

Short answer: A lot.

Long answer: The burn to Jool from outside Kerbin's SOI is frighteningly close to the burn in LKO. So a lot of the fuel you burn to get out of Kerbin's SOI is wasted.

Longer answer: You can't both have fuel constraints AND terribly low TWR. The whole point of having low TWR is because fuel is cheap so you can bring as much as you want. I'd find a way to add 30-50% more fuel and make that burn 3 hours instead of 2.

EDIT: actually I wouldn't do that. Like @Plusck I don't do these kinds of burns. What I'd do is find a way to add about 10x the engine power to make the burn 10 minutes or so instead of 2 hours.

Edited April 5, 2016 by 5thHorseman

[March Unto Torment]

Multiplying my TWR by ten is... not an option I'm loathe to reveal the exact nature of the mission, but I've more or less already stacked on all the engines I have room for. I could maybe double my engine power, but that'd add a lot of weight (and hence cut out a lot of valuable dV) and not make a huge dent in my burn time.

Anyway, bringing more fuel is probably my best bet. It means another set of fuel tanks to add on to the ship and another three fuelling runs or so to KEO (I'm bringing my tanks up empty for weight reasons), but it'll probably let me get the extra 2,270m/s on (going by your numbers, it'd be more efficient than the KEO burn and the dV for exiting Kerbin's SOI plus burning to Jool is only about 5,000m/s in total - does that look right to you?).

[Streetwind]

59 minutes ago, 5thHorseman said:

One other option is to do it the way that's wrong but that most new people do: Burn just enough to get out of Kerbin's SOI (I suggest using a Mun slingshot for this, it'll save you some of the dV you're wasting doing it this way) and then once in Sun orbit plot and execute the Jool transfer.

I'm wondering, in the scenario where a Mun gravity assist is a thing:

How much burning could you do while approaching and passing the Mun, for some Oberth help? Around twenty minutes should be possible, no? Then instead of just creeping into solar orbit and doing the whole transfer there, you'd have a shortened burn out there. And because of the poor efficiency of burning in solar orbit, this should theoretically decrease the cost (and burn time) there significantly...

It would result in three burns:
- One periapsis kick towards a close Mun pass
- One propulsively assisted Mun gravity slingshot
- One burn in solar orbit

[March Unto Torment]

5 minutes ago, Streetwind said:

I'm wondering, in the scenario where a Mun gravity assist is a thing:

My difficulty with gravity assists is that I find them very unpredictable, and I have yet to find a decent way of figuring out how much dV they use. Usually, I'd compensate for this just by packing in spare dV, but due to the nature of this mission, I'm trying to operate on as lean a dV budget as I can. Do you happen to know any way of calculating an effective gravity assist dV?

[Superfluous J]

5 minutes ago, March Unto Torment said:

My difficulty with gravity assists is that I find them very unpredictable, and I have yet to find a decent way of figuring out how much dV they use. Usually, I'd compensate for this just by packing in spare dV, but due to the nature of this mission, I'm trying to operate on as lean a dV budget as I can. Do you happen to know any way of calculating an effective gravity assist dV?

Calculating? No. But you can plot maneuver nodes and just add them up. Of course, any long burns from those nodes will get less and less accurate, with your low TWR.

Were I to do this, I'd just aim to get my Mun slingshot to throw me out of the Kerbin system prograde and a bit radially away from the Sun. I would just be happy for the free dV it provided, however little or much that ended up being.

[Streetwind]

It really depends strongly on your starting conditions, and the angle you hit it at. So I can't give you any "always true" figure. I can, however, give you a way to figure things out in your specific case.

From your mothership's parking orbit, plot a maneuver node. Find a Mun encounter, and tune it so that you pass behind it (from Kerbin's point of view). This will accelerate you. Make sure the pass is a nice low periapsis (sub 20 km), and tune your maneuver node so that you're just barely escaping Kerbin's SoI after the Mun pass. It may take some playing with the node and moving it around to get the lowest dV cost possible for this.

Once you have that dV cost, rightclick your maneuver node and add a couple orbits to it - whatever is enough to make you miss the Mun completely. You should now have a projected orbit within Kerbin's SoI.

Finally, add dV to the node until you again barely escape the SoI. The amount you just added - the difference between escaping with the Mun and escaping without - is what your Mun slingshot saves you.

 

However, keep in mind that you can save more than that with a propulsively assisted slingshot. What I was suggesting is, since you are planning a long burn in solar orbit, you might as well try and get part of that burn done while you are in close proximity of the Mun. Because there, you get the Oberth effect to help you, whereas in solar orbit, you don't. You potentially stand to gain far larger savings through doing this than the normal Mun slingshot would give you.

That said, I've never done such a maneuver myself. It would take careful pre-planning to see if (and how) it works. Start with the same minimum-dV-to-Kerbin-escape-via-Mun-slingshot maneuver node you've made above. Now put a second maneuver node onto your solar orbit trajectory and plot a Jool intercept - if you are launching within the proper window, you should be able to get it directly from there. Note how much dV it costs, and note the rough shape of the orbit.

Now delete the second node. Instead, make a node at your Mun periapsis (which, as mentioned before, should be nice and low). Add dV until you have about a 15 to 20 minute burn. Now, it's important to make sure that the solar orbit you get out of that is shaped roughly similar to the complete Jool transfer orbit you planned earlier, even though this one doesn't go all the way to Jool. Apoapsis should still point roughly in the right direction, and such. The Mun is also a good place to make an inclination change, If you have to - you can get is practically for free just by adjusting the flyby trajectory.

Finally, make a third maneuver node, this one again just outside Kerbin's SoI. That's the node you'll use to complete your transfer with. Add dV until you have your encounter. Now add up your dV here and that of you Mun node, and compare it to the total cost of the single solar orbit maneuver you planned at the start. You should see a noticable cost reduction. Theoretically.  

I'd wager that this would be one of the most intricately pre-planned maneuvers you've ever done! And thus I'm sure it will thrill you to know that as soon as you made the first burn, you can throw node 2 and 3 into the trash! This is unfortunately a result of the imprecision in executing your burn manually; you'll never be able to hit the second node exactly where you planned it to be. So you must delete and remake it. But it should be easier this time, since you know roughly how you set it up the first time and how the resulting orbit looked. Make the second burn during your Mun pass, and after that you can remake your third node properly.

 

EDIT: I should probably add that all of this works only when you hit the Mun in such a way that it kicks you out of Kerbin's SoI in the prograde direction. You can't just slingshot in any old direction. It has to be the same one you'd push your orbit into when making a traditional Hohman transfer burn, and as Jool is an outer planet, this means ejecting prograde with respect to Kerbin's solar orbit.

Because your ship is geostationary, you have a long orbital period. You may find that you have difficulties hitting the Mun in exactly the right way to get this trajectory, even when shifting your maneuver node around. The reason for that is that your ship will simply be on the wrong side of its orbit at the proper time, and take too long to swing around. In that case, if you have time left until the transfer window comes around, you can try and gently "phase your ships's orbit" by deviating slightly from geostationary altitude - either increasing or decreasing your orbital period with just a little burn. Over many consecutive orbits, even small changes of 5-10 minutes per orbit can add up into a totally different vessel position when the time for the transfer rolls around, letting you hit the Mun slingshot with much more precision.

Edited April 5, 2016 by Streetwind

[Plusck]

1 hour ago, March Unto Torment said:

My difficulty with gravity assists is that I find them very unpredictable, and I have yet to find a decent way of figuring out how much dV they use. Usually, I'd compensate for this just by packing in spare dV, but due to the nature of this mission, I'm trying to operate on as lean a dV budget as I can. Do you happen to know any way of calculating an effective gravity assist dV?

Calculating - like 5thHorseman and Streetwind have said, I don't think there's any easy way to do that.

However, it's easy enough to SWAG:

- the Mun's orbital velocity is about 540 m/s (iirc). So if you approach slowly from LKO, get dragged around 270° and ejected prograde, you're adding 540 m/s horizontal (with respect to Kerbin) velocity to your original SOI-entry velocity (since you'll leave the Mun's SOI at exactly the same speed as you entered it). To get such a large, 270° swing around the Mun you'd need to be going pretty slowly when you enter the Mun's SOI, and from what we all see when plotting our first Mun intercepts, that 270° swing tends to be when your Ap is very close to the Mun's own orbit;

- according to this excellent calculator, a burn from LKO to the Mun ends up with orbital velocity of about 170 m/s when you reach the Mun. Therefore the Mun's SOI catches up with you at an absolute maximum of about (540-170) = 370 m/s. You must therefore enter the Mun's SOI at a velocity of, say, 400 m/s (taking the vertical component of your velocity into account), and if you exit precisely prograde you'll leave it with 400+540= 940 m/s. Escape velocity is about 750 m/s at the Mun's altitude, so you're basically at escape+180 m/s, which is more or less right to get to Duna or Eve.

- again according to that excellent calculator, a transfer from KEO to the Mun has an Ap velocity of about 400m/s, so the Mun's SOI catches up with you at about 140 m/s. If you go 270° around the Mun, you'll leave prograde with an orbital velocity relative to Kerbin of about 720-740 m/s. Not quite escape...

- the trouble with Jool is that you're wanting to end up with around Kerbin escape+980 m/s. Therefore around 1730 m/s on leaving the Mun's SOI, therefore 1730-540= 1190 m/s at the edge of the Mun's SOI.

So therefore, if you are transferring up from KEO to the Mun for an unpowered gravity assist, and going 270° around the Mun to eject prograde, you need to find another 1050 m/s from somewhere to get to Jool.

 

* DISCLAIMER: all or some of the above information might be wrong. I don't have KSP open and I don't have anything in KEO that I could easily use to test, so the figures are based off that calculator I was talking about, using figures largely taken from memory. You may misremember reality, even as it is happening, as being different from my pure unsullied memories of it.

Also please note that if you do less than a 270° swing around the Mun, you'll get increasingly less of an assist. It'll always give you a boost if you go the right way, but less than 400 m/s or so from a 270° swing.

[KerikBalm]

You should only need about 2,000 m/s to get to Jool from LKO. If you are in KEO, you should do a bi eliptic transfer, drop your PE down to just about kerbin's orbit, then burn at PE.

Since you're starting in an eliptic orbit, not a circular low orbit, you need even less than 2,300 m/s. Its not like you're coming from Mun, but close... lets deduct ... 650 m/s from the 2000 needed from a 70x70km circular orbit

Now your required burn is 1350 instead of 4,131... down to 1/3 the amount.

Now you can easily throw away half the fuel(even more than half, because dV does not scale linearly with fuel amount, although linear approximations work for small dV values... 4100 vs 1800 is not small) on your transfer stage for a much better TWR (considering the burn to lower your PE, you'll need about 1,800m/s total)... you could also add more engines, as you can afford the extra mass because you've reduced the dV requirement.

If its too late and you've already attached the under powered overfueled transfer stage, then still do the bi eliptic burn.

1350 instead of 4100 means a 39 minute burn instead of a 2 hour burn... still not great at all... but what do you expect when you design a ship with barely better than a half meter per second burn time, knowing you'll need burns in the thousands of m/s?

You can perapsis kick the first burn to get your apoapsis past minmus, and your second burn will only need to be about 1050...

Down to a 30 minute burn... start burning halfway between prograde and the manuever node 15 minutes before the maneuver node... and keep burning... you'll have thousands of dV to spare this way.

[March Unto Torment]

13 minutes ago, KerikBalm said:

Now your required burn is 1350 instead of 4,131... down to 1/3 the amount.

 

Well, now this changes everything. I've actually more or less cut my fuel weight by 2/3 (not sure why it scales so linearly, although it's probably because - even when fully-fuelled - the craft was only originally less than 50% fuel), and I still have 2,000m/s left, which gives me plenty for correction burns once on a transfer trajectory. In doing this, I've increased my TWR to around the 0.1 end of the spectrum - still not fantastic, but combined with the far lower dV of the burn, it should get my burn time down to around - as you said - 30 minutes rather than two hours. Ideally, I'll be able to break my burn down even further, by splitting it into one burn to edge of SOI and a second to Jool.

This is all assuming that you're right, of course. But if you are, this whole setup just got way more practical.

[KerikBalm]

Don't forget the dV you need to lower your perapsis to about 70km

Since orbits are reversable:

http://wiki.kerbalspaceprogram.com/wiki/KEO

If it was a 434 m/s burn to circularize from a pe of 70K to KEO, it will take 434 m/s to drop pe down to 70km from KEO.

Then that means at Pe, you'll be moving 676 m/s faster than what you need for a 70x70km orbit... thats the same as if you burned the first 676 m/s of your transfer burn to jool

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

And as you'll see there, you need less than 2,000 m/s to get to jool from 70x70km.

But be careful! the timing on this is not easy.

You can't lower your PE from anypoint at KEO, you have to lower it so that your PE is where you'd place the maneuver node if you were in a 70x70km orbit, and you need to arrive at PE close to the middle of the transfer window (not such a problem from KEO... a much more complex problem if you were at minmus trying to do this)

https://alexmoon.github.io/ksp/#/Kerbin/70/Jool/100/false/optimalPlaneChange/false/1/1

This is the dV cost with a mid course plane change as opposed to the more complex "ballistic" trajctory where you include a normal/antinormal component to your ejection burn/start the trasnfer from an inclined orbit

1959m/s ejection burn, 58 m/s plane change, 2017 m/s total if you get it juuusssssssst right (keep in mind, you can subtract 676 from this if you lower your Pe from the correct point, and arrive at PE at the transfer window)... Year 2, day 242 departure... I just clicked a spot in the year 1 transfer window... 1948 m/s ejection, 125 m/s plane change... 2073 total...

You're still going to have a long burn that will result in inefficiency... don't cut it too close to the "ideal" dV budget, because the situation of lowering your PE at the right place and doing a burn with low acceleration is not ideal

I'm assuming your transfer stage is just to get you to Jool, not to handle the capture (I recommend tylo or laythe gravity capture, then laythe aerobraking, but that depends on what you want to do after arriving at Jool).

A propulsive capure can cost a bit (ignore what the transfer window calculator says... thats to capture into a 100km orbit around Jool, which I assure you, you don't want to do, as that is within its atmosphere... I assume you want to visit is moons, not dispose of the station in a gas giant)

 

 

[Plusck]

On 05/04/2016 at 4:32 PM, KerikBalm said:

You should only need about 2,000 m/s to get to Jool from LKO. If you are in KEO, you should do a bi eliptic transfer, drop your PE down to just about kerbin's orbit, then burn at PE.

Since you're starting in an eliptic orbit, not a circular low orbit, you need even less than 2,300 m/s. Its not like you're coming from Mun, but close... lets deduct ... 650 m/s from the 2000 needed from a 70x70km circular orbit

Now your required burn is 1350 instead of 4,131... down to 1/3 the amount.

Now you can easily throw away half the fuel(even more than half, because dV does not scale linearly with fuel amount, although linear approximations work for small dV values... 4100 vs 1800 is not small) on your transfer stage for a much better TWR (considering the burn to lower your PE, you'll need about 1,800m/s total)... you could also add more engines, as you can afford the extra mass because you've reduced the dV requirement.

If its too late and you've already attached the under powered overfueled transfer stage, then still do the bi eliptic burn.

1350 instead of 4100 means a 39 minute burn instead of a 2 hour burn... still not great at all... but what do you expect when you design a ship with barely better than a half meter per second burn time, knowing you'll need burns in the thousands of m/s?

You can perapsis kick the first burn to get your apoapsis past minmus, and your second burn will only need to be about 1050...

Down to a 30 minute burn... start burning halfway between prograde and the manuever node 15 minutes before the maneuver node... and keep burning... you'll have thousands of dV to spare this way.

Not quite.

The burn from KEO down to LKO is about 450 m/s. Total cost is therefore about 1800 m/s (as opposed to 2100 m/s direct from KEO), so there is some saving.

However that doesn't help if you have a very low TWR: on an eliptic orbit like that there is simply not enough time at Pe to do more than a couple of minutes of burn. 39 minutes is quite simply impossible - you'd be back out well past KEO by the time you finished, so that 450 m/s cost of dropping to LKO would have been a pure waste. Actual dv cost for that 39 minute burn would be back up to over 2000 (and would last a lot longer than 39 minutes).

And there is no way you could start a burn 15 minutes before Pe on an KEOxLKO orbit. Your burn target marker wouldn't even be pointing directly at the centre of Kerbin, it would be pointing through the atmosphere on the far side of Kerbin when you started the burn. You wouldn't risk dropping back into the atmosphere - you'd be certain to power straight into the ground.

[KerikBalm]

Did you not read all of what you quoted?

"Now your required burn is 1350 instead of 4,131... down to 1/3 the amount.

Now you can easily throw away half the fuel(even more than half, because dV does not scale linearly with fuel amount, although linear approximations work for small dV values... 4100 vs 1800 is not small) on your transfer stage for a much better TWR (considering the burn to lower your PE, you'll need about 1,800m/s total)... you could also add more engines, as you can afford the extra mass because you've reduced the dV requirement."

Yes, the acceleration is still horrible, and while the final burn only needs to be 1350, the time to execute it is much worse. Considering he could have dV out the wazoo, he could start say... 7.5 minutes ahead of the node instead of 15 minutes, and take the penalty on the burn after the node... burning more to make up for it.

Also, note that while 1350 is roughly 1/3 the amount of 4131... the LFO engine do not have 1/3 the amount of Isp as the nukes

1350/4131 = 0.327

340/800 = 0.425

KR-2L vs nukes... 1 KR-2L has the mass of 3 nukes, but over 11x the thrust

Personally I use chemical engines to help eject my stuff from LKO, which then continues on with nukes.

This was my Jool mission:

due to the docking port connections, I had to limit acceleration to 6 m/s/s even though it could do 1G in theory.

I only did 1 perapsis kick... it could have been more efficient if I carried more LF, and less OX, and PE kicked with nukes, but I was not so patient, and used the KR-2L for the PE kick too

Even after ditching the KR-2L booster, I was still using LFO engines to help get to the required velocity.

Those two aerospikes put out 360kn, while the remaining 4 nukes only put out 240 kn

Just FYI, the booster was reusable, after detaching, I turned it retrograde, and burned its nukes:

As you can see, I stopped it from escaping, and actually left too much LF in there, I don't need 950 m/s to raise the PE out for aerobraking, and then raise the PE again to stabilize the orbit (with that much dV, I may just try to get to one of the moons of kerbin to get refueled from ISRU there, rather than wait in LKO for a refeuling shipment from one of the moons)

[Plusck]

11 hours ago, KerikBalm said:

Did you not read all of what you quoted?

"Now your required burn is 1350 instead of 4,131... down to 1/3 the amount.

Now you can easily throw away half the fuel(even more than half, because dV does not scale linearly with fuel amount, although linear approximations work for small dV values... 4100 vs 1800 is not small) on your transfer stage for a much better TWR (considering the burn to lower your PE, you'll need about 1,800m/s total)... you could also add more engines, as you can afford the extra mass because you've reduced the dV requirement."

Yes, the acceleration is still horrible, and while the final burn only needs to be 1350, the time to execute it is much worse. Considering he could have dV out the wazoo, he could start say... 7.5 minutes ahead of the node instead of 15 minutes, and take the penalty on the burn after the node... burning more to make up for it.

Also, note that while 1350 is roughly 1/3 the amount of 4131... the LFO engine do not have 1/3 the amount of Isp as the nukes

1350/4131 = 0.327

340/800 = 0.425

KR-2L vs nukes... 1 KR-2L has the mass of 3 nukes, but over 11x the thrust

Personally I use chemical engines to help eject my stuff from LKO, which then continues on with nukes.

This was my Jool mission:

 

I only did 1 perapsis kick... it could have been more efficient if I carried more LF, and less OX, and PE kicked with nukes, but I was not so patient, and used the KR-2L for the PE kick too

 

Those two aerospikes put out 360kn, while the remaining 4 nukes only put out 240 kn

As you can see, I stopped it from escaping, and actually left too much LF in there, I don't need 950 m/s to raise the PE out for aerobraking, and then raise the PE again to stabilize the orbit (with that much dV, I may just try to get to one of the moons of kerbin to get refueled from ISRU there, rather than wait in LKO for a refeuling shipment from one of the moons)

Yes but that "4100 m/s" number corresponds to nothing at all. Making up numbers isn't a good way to calculate efficiencies.

Jool burn from KEO is about 2100 m/s. NOT 4100 m/s. There is nowhere in the Kerbol system that takes a 4000 m/s burn to get to from anywhere else (edit: with the exception of Moho to Jool/Eeloo, of course).

Did you not read my post?

Edited April 7, 2016 by Plusck

[KerikBalm]

Did you not read the OP?

"just the stage I'm using to get there from KEO, which has almost exactly enough dV to get to Jool from KEO (4,131m/s) - is sort of ridiculous."

Personally i haven't calculated the burn from KEO, but the OP says he's trying to get to Jool and his plotted burn is required 4,131 m/s.

I'm pointing out that from KEO, he should be able to get there from KEO for 1,800 m/s, with the largest burn being only 1,350, which significantly reduces the burn time.

 

You said "The burn from KEO down to LKO is about 450 m/s. Total cost is therefore about 1800 m/s" ir reply to my post in which I said: "considering the burn to lower your PE, you'll need about 1,800m/s total"... so if you quote me and "correct" me by basically saying the same thing, I must conclude you didn't really read all of what you quoted.

[Plusck]

And to @March Unto Torment - I don't know where you got that "3800 m/s" figure from but it certainly looks like you're misreading the delta-v maps.

In that thread I linked to earlier, the most efficient burn altitude for transfer to Jool is 371km. This means that from KEO it'll be less than optimal, but it'll never be anywhere near 3800 m/s.

What you need, to get to Jool, is escape velocity +980 m/s from LKO. This is because if you go that fast from LKO (i.e. 2300 m/s orbital velocity + 950 escape + 980 to get to Jool = 4230 m/s total orbital velocity on leaving LKO), you don't lose too much before you get to the edge of the SOI, and so it gets added to Kerbin's own orbital velocity out in interplanetary space. If you are just coasting in space near to Kerbin's SOI, you need to burn about 2600 m/s to get to Jool.

So wherever you start in the Kerbin system, all you need to do is to leave Kerbin's SOI with a velocity of about 2600-2700 m/s.

So starting from Minmus, for example, you are quite near the SOI edge and it doesn't take much of a burn to get there, but you still need to find that 2600 m/s added velocity from somewhere. That's why it's more expensive to go to Jool from Minmus than from a lower altitude.

On 05/04/2016 at 3:49 PM, Plusck said:

...

- the trouble with Jool is that you're wanting to end up with around Kerbin escape+980 m/s. Therefore around 1730 m/s on leaving the Mun's SOI, therefore 1730-540= 1190 m/s at the edge of the Mun's SOI.

So therefore, if you are transferring up from KEO to the Mun for an unpowered gravity assist, and going 270° around the Mun to eject prograde, you need to find another 1050 m/s from somewhere to get to Jool.

...

This figure of 1730 isn't correct - from the Mun's altitude, you need to burn 2500 m/s to get an orbital velocity of 3000 m/s to get to Jool.

Still a far cry from 3800 m/s.

And as you get nearer to 371 km altitude (that "optimal" altitude for a Jool burn) it'll decrease.

[March Unto Torment]

http://13375.de/KSPDeltaVMap/

I was using this dV map, which was indicating wildly out-of-proportion dV numbers (3,860 for KEO to LJO). Reading everything in this thread, that definitely seemed wrong.

So, I tested it out myself with a manoeuvre node and it turns out that to get to Kerbin's SOI edge (KSE from here on out), it only takes me 390m/s from KEO - not the 2,270m/s that my dV map was indicating (for which matter, the love is wrong with that dV map?!). This stands to reason, as it's 660m/s to raise AP to KEO, and 950-660=390.

So, from that, I infer from Swash's dV map (which seems much more reliable) that I need 1410 to make the transfer from SOI edge to Jool elliptical, which seems reasonable to me. That brings me up to a round 1,800 for KEO to something resembling an orbit around Jool.

From there, I need to make a transfer to... an orbit around Jool. I'm keeping the mothership in orbit around Jool while using a separate tug to move my landers around the system for operations, but I'm not sure where would be the most dV-efficient place to put my mothership (I'm thinking around 20,000km, putting me just inside Laythe orbit; it's staying there indefinitely and won't be making the return to Kerbin, due to my 'unusual' mission profile).

Anyone else have any other suggestions for where to plant the mothership? And does anyone have any idea how much dV it would take to get from Jool capture to that orbit?

EDIT: The forum autocorrect to 'love' is hilarious  Also, I'm not quite sure what you're saying about orbital velocity - exactly how much dV do I lose by starting with my PE at KEO?

Edited April 7, 2016 by March Unto Torment

[Plusck]

1 hour ago, KerikBalm said:

Did you not read the OP?

"just the stage I'm using to get there from KEO, which has almost exactly enough dV to get to Jool from KEO (4,131m/s) - is sort of ridiculous."

Personally i haven't calculated the burn from KEO, but the OP says he's trying to get to Jool and his plotted burn is required 4,131 m/s.

I'm pointing out that from KEO, he should be able to get there from KEO for 1,800 m/s, with the largest burn being only 1,350, which significantly reduces the burn time.

 

You said "The burn from KEO down to LKO is about 450 m/s. Total cost is therefore about 1800 m/s" ir reply to my post in which I said: "considering the burn to lower your PE, you'll need about 1,800m/s total"... so if you quote me and "correct" me by basically saying the same thing, I must conclude you didn't really read all of what you quoted.

No need to do selective quoting.

I said, in answer to you "The burn from KEO down to LKO is about 450 m/s. Total cost is therefore about 1800 m/s (as opposed to 2100 m/s direct from KEO)"

But maybe I should have added an "indeed". Maybe I should also highlight that for you:

"The burn from KEO down to LKO is about 450 m/s. Total cost is indeed therefore about 1800 m/s (as opposed to 2100 m/s direct from KEO)"

The second problem is you're totally ignoring the fact that with an ultra-low TWR, it is pointless getting into an orbit where you have 10 minutes max to do your burn at Oberth-effect speeds.

And that figure of 4131 m/s is the OP's ship. And you should know, without even thinking about it, that that 3800 m/s burn is impossibly large for Jool.

[Plusck]

1 hour ago, March Unto Torment said:

So, from that, I infer from Swash's dV map (which seems much more reliable) that I need 1410 to make the transfer from SOI edge to Jool elliptical, which seems reasonable to me. That brings me up to a round 1,800 for KEO to something resembling an orbit around Jool.

 

hmm - I'm still not sure you're reading the dv map correctly.

As @5thHorseman said right at the beginning of this thread, the burn from just outside Kerbin's SOI is horrible. In fact, it's about 2600-2700 m/s.

The figures in the standard dv maps for interplanetary transfers are "from low orbit, in addition to the amount to get to SOI edge". So Jool is +950+980 from LKO. It will certainly be slightly more from KEO - and I found it to be about 2100 m/s. See also, for a good planner, https://alexmoon.github.io/ksp/#/Kerbin/3400/Jool/100/true/ballistic/false/2/400

1 hour ago, March Unto Torment said:

From there, I need to make a transfer to... an orbit around Jool. I'm keeping the mothership in orbit around Jool while using a separate tug to move my landers around the system for operations, but I'm not sure where would be the most dV-efficient place to put my mothership (I'm thinking around 20,000km, putting me just inside Laythe orbit; it's staying there indefinitely and won't be making the return to Kerbin, due to my 'unusual' mission profile).

You have two real options for arrival at Jool: gravity assist at Tylo or at Laythe.

The problem with arriving at Laythe is that going from there to the outer moons is expensive. In addition, a zero-cost gravity assist will put you into a sub-Laythe to near-Vall orbit, which must be corrected to avoid crashing into one or other moon.

A gravity assist from Tylo will put you into a Vall-to-Bop orbit. Chances of a collision within dozens of days are much lower, and it won't take much to use that to get to all of the moons without too much difficulty. So personally I'd aim for Tylo, swing by, and only actually enter Tylo orbit just before you want to leave the system (since burning to Kerbin from Tylo is much more efficient than burning from Tylo orbit out to Jool orbit then from there to Kerbin, for exactly the same reasons as you don't want to burn out to Kerbin SOI edge and then burn to Jool in the first place).

 

 

edit: re "orbital velocity"

This is actually the thing that determines how much of a burn you need to go anywhere.

In LKO, orbital velocity is about 2300 m/s. If you increase this by 900 m/s (so at Pe, orbital velocity is 3200 m/s) you will raise your Ap to near Minmus's orbit.

If you increase that by another 50 m/s or so, you escape Kerbin's SOI. However, at the point where you leave Kerbin's SOI, your orbital velocity will have dropped to nearly zero (you can also think of it as Kerbin's gravity dragging you back).

However, if you increase that by another 1000 m/s (so orbital velocity as you leave Pe at 70km will be 4200 m/s), you will reach the SOI edge much more quickly, Kerbin's gravity will have much less time to slow you, and your orbital velocity relative to Kerbin will be much, much higher than that "additional" 1000 m/s. In fact, it will be around 2600 m/s at the SOI edge, so you enter interplanetary space with Kerbin's own orbital velocity +2600 m/s, which gets you to Jool.

 

Edited April 7, 2016 by Plusck