Low Power - Long Distance

[NewtSoup]

I once made an ion craft in a sandbox game and spun it out to Duna but it was able to do it in a single burn.

When I watch videos on youtube I often see people doing multiple burns at periapsis to take advantage of the oberth effect, which I get.

What I don't know though is how do you calculate what burn you need to do the transfers?

Obviously you can't set up a single node ( or maybe you do ) for a single burn and surely after say 5 burns the planet your orbiting will be a day or 6 ahead in it's orbit so how do you know if you're still on target?

The only video I've seen that even briefly covered how to do it was one of Matt Lowne's  but it was very brief, I didn't quite get it the first time around and now I can't find which video it was.

Can someone explain it to me or direct me to a guide on the subject?

[DeltaDizzy]

7 minutes ago, NewtSoup said:

I once made an ion craft in a sandbox game and spun it out to Duna but it was able to do it in a single burn.

When I watch videos on youtube I often see people doing multiple burns at periapsis to take advantage of the oberth effect, which I get.

What I don't know though is how do you calculate what burn you need to do the transfers?

Obviously you can't set up a single node ( or maybe you do ) for a single burn and surely after say 5 burns the planet your orbiting will be a day or 6 ahead in it's orbit so how do you know if you're still on target?

The only video I've seen that even briefly covered how to do it was one of Matt Lowne's  but it was very brief, I didn't quite get it the first time around and now I can't find which video it was.

Can someone explain it to me or direct me to a guide on the subject?

You mean "How do you figure out how many burns you need?" I usually just eyeball it. By that I mean I burn at Periapsis(Pe), and if im still burning too far* past Pe I just go around again.

*too far is basically when I feel like it's gone on too long pas Pe

Edited May 29, 2018 by DeltaDizzy

[NewtSoup]

1 minute ago, DeltaDizzy said:

You mean "How do you figure out how many burns you need?" I usually just eyeball it.

Well that and how do you know you're going to get an encounter on the last burn?  When I took the ion craft to duna I simply put it into an orbit around kerbol first and then created a node on that orbit to get an encounter with Duna, I had no idea of transfer windows or anything I just went out and did it. It was probably hugely inefficient but I gave that ship 19k of DV :D.. It was basically 2 ion engines a load of RTGs and fuel tanks and a command chair.

hold on.. I have a picture somewhere.

Basically I brute forced it.  Now I want to know how to do it "properly"

[bewing]

Well, there's the concept of finding the window, and then trusting the window. The easiest way to find the window is just to look it up in an online transfer planner. To find windows without a transfer planner, you basically need a test ship -- do you know how to do that part?

But if you know that your transfer window is in 2 days and 3 hours, and you are at your Pe -- then you arrange your next to last burn so that the resulting period is 2 days and 3 hours. When you get to the final burn, you just burn until you have either a projected encounter, or until your Ap is just a little above the target planet's orbit and your "Closest Approach" is pretty close. Then you do a midcourse correction to get the encounter really nice.

So, which parts of that do you want more clarity on?

 

[Superfluous J]

Periapsis kicking is similar to anything in KSP: You start at where you want to go and work backwards from there.

So, what you want is to eject to Duna from the periapsis of a highly elliptical orbit, at day (say) 123.

So in order to do that, you need to burn up into that orbit on day 123-x, where x is the period of that orbit. There are several things to consider, such as where that periapsis is and what is that orbit's period, and those are generally answered by doing a few tests or at least drawing out maneuver nodes.

There are 2 schools of thought - at least that I've had - about the eccentricity of the orbit. Either you should stop short of Mun so you won't have to worry about it, or you should burn up pretty high - higher than Minmus but still within Kerbin's SOI - and trust your patched conics to not fail you so you're sure you're not going to get a Mun encounter. The first one has the benefit of no guaranteed Mun encounter and a the detriment that your final burn will be a little bit longer - and less precise and efficient. The second one has the opposite benefit and problem.

To get the time, just create a maneuver node and drag it out, and see how long until the next periapsis. For sub-Mun orbits I don't think it can get much over a day or two. For trans-Minmus orbits it can be well over a "munth" - which is only 6 days. But it can pretty easily be 15 or so days.

To get the location, create another maneuver node to actually transfer to Duna. You'll have to of course be X days or more before the desired transfer and will have to shift the node forward in time that many days. But draw your transfer node until you get your encounter with Duna, and then leave it there as a marker for where to place your actual burn up into your elliptical orbit.

And from there it's pretty straightforward. Burn up into the elliptical orbit, then re-do that ejection node and burn it.

Kicking more than once just adds the wrinkle that you have to plan more than one ellpitical orbit. But it's exactly the same process just iterated a few more times.

[Snark]

18 hours ago, NewtSoup said:

What I don't know though is how do you calculate what burn you need to do the transfers?

Obviously you can't set up a single node ( or maybe you do ) for a single burn and surely after say 5 burns the planet your orbiting will be a day or 6 ahead in it's orbit so how do you know if you're still on target?

The only video I've seen that even briefly covered how to do it was one of Matt Lowne's  but it was very brief, I didn't quite get it the first time around and now I can't find which video it was.

Can someone explain it to me or direct me to a guide on the subject?

So there are a few different problems mixed up and tangled together here, which can muddy the waters a bit.  Let's sort them out and address one by one.

The first question is:

How do you calculate how much total dV you need for a transfer burn?

Notice the use of the word "total" here.  This is deliberately ignoring any question of whether you do multiple burns with periapsis kicking, or whatever.  It's simply saying, "to get from Planet A to Planet B, how much dV is needed?"

There are various ways to do this.  Here are a few possibilities:

1. Use a transfer-burn calculator.  There are a lot of them out there.  My personal favorite is http://ksp.olex.biz, just because it's so nicely simple to use and has a nice graphical illustration of where/how to burn.  You just tell it your origin planet, destination planet, and the height of your parking orbit above the origin, and it tells you how much burn you need, and in what direction, and when the planets are in what relative position to each other.
2. Just use a maneuver node.  Plop a node down, tinker with it to get an intercept to your target.  There you go.  That's the burn you need.
3. Do the math yourself,  using the vis-viva equation.  Involves some button-mashing on a calculator (or setting up a spreadsheet), but it works pretty well and the KSP wiki provides all the useful numbers such as planets' gravitational parameters and orbit sizes.

Okay, so that's one problem solved.  "I'm in 100 km orbit over Kerbin, and I want to go to Duna.  How much dV do I need?"  Answer, 1043 m/s.  There, done.

Okay, on to the next question:

How many burns do I need?

After you work out the total dV needed, figure out your total burn time.  The simplistic approach is to just take the dV, multiply by ship mass in tons, and divide by engine thrust in kilonewtons; that's how many seconds.  This is not strictly accurate, since it fails to account for your ship's increasing acceleration as its mass goes down as it burns fuel, but for a high-Isp engine like ions, it'll be pretty close.  If you want more accurate than that, you can do the math based on fuel consumption rate, or use a mod like BetterBurnTime to do the math for you.

Once you know your total burn time, it's a matter of carving up the burn into "reasonable" chunks.  I like to keep any individual burn under 3 minutes, myself, to avoid screwing up the navigation too much.  So, if I work out that I've got 8 minutes of total burn time, say, then I'll split that up and do three burns of 160 seconds each or thereabouts.

However, there's an important caveat about burn times-- planetary escape velocity places a ceiling on how much dV you can gain while "kicking" before you have to do your final escape burn.  More on this below.

 

How do you keep track of your burns with maneuver nodes if you're doing multiple burns with periapsis kicking?

I dunno how other folks do it, but here's what I do-- works pretty well for me, isn't too inconvenient:

1. Set up a maneuver node as if I were going to do the whole thing in one burn.  e.g. "Okay, here I am in 100 km LKO, I want to go to Duna, I'll plop down my 1043 m/s maneuver node riiiiight... there."
2. Decide how I'm going to split my burn up.  "Okay, let's see, I need 8 minutes burn time.  I'll do this in three burns of 160 seconds each."
3. Do the initial burn at the maneuver node, leaving the maneuver incomplete.  "Okay, here comes the maneuver node.  My first burn is 160 seconds.  Half of that is 80 seconds.  I'll start burning when it's T minus 80 seconds, and stop burning when it's T plus 80 seconds."
4. As soon as the initial burn is complete, note how many m/s of the burn remains.
5. Delete the maneuver node.  Then plop down a new one in the same spot (it's easy to locate, since it'll be right at Pe of your new orbit), drag its handle until it has as many m/s as I noted in step 4, then switch to map view and fine-tune the burn until I once again have an intercept on my interplanetary target.  (This only takes a few seconds, since plop-it-down-at-Pe-with-about-the-right-dV gets pretty darn close and only a small adjustment is needed.)
6. Repeat steps 3 through 5 until I've done all but the final (escaping) burn.
7. On the last burn, just burn whatever dV is remaining until I get within about 100 m/s of completing the burn.  I'm now on an escape trajectory from my origin planet.
8. Delete the maneuver node, plop down a new one a few minutes ahead, and then fine-tune that one to get exactly the encounter I want.  Do that burn the usual way (it'll be a short one, since it's only a small amount of dV).

Works great, it's nice and accurate.

 

Important caveat about the "ceiling" for periapsis kicking:

If you're already in LKO... there's only so much dV you can add with periapsis kicking, because beyond a certain point you'll exceed Kerbin's escape velocity and then you can't do any more kicking.  Kerbin escape velocity is only around 950ish m/s above orbital velocity, if you're in LKO.  So you can't add more than that before you have to do your final, continuous escape burn.  So, if you work out that you need to do a burn much bigger than that to get to your target-- say, 3000 m/s-- you'll only be able to use periapsis kicking for the first few hundred m/s, and your final "escape" burn will need to be a giant one, potentially thousands of meters per second.  If your TWR is really low, that may mean that your final "escape" burn will need to be really really long, and you'll lose a giant chunk of Oberth benefit, but that's how the cookie crumbles.  No way around it.

The way I generally handle this sort of scenario, when it comes up:

1. First I do about 700 m/s or so of periapsis kicking.  Every bit helps. 
2. For the (giant) escape burn, set up a maneuver node.
3. Start the burn no more than about 3 minutes before periapsis.  Aim about halfway between and .  Keep it at that halfway-between mark until the two merge, then just set SAS to follow .
4. Burn until I have about 400 m/s of dV remaining on the maneuver node.
5. At this point I assume that my node is off due to the long burn.  So I delete the burn and plop down a new node 20 or 30 minutes ahead, and fine-tune that for an encounter.
6. Do that burn the usual way.  By that point we're far enough from Kerbin that the somewhat-long remaining burn time doesn't matter much.

Finally, there's this question:

18 hours ago, NewtSoup said:

Obviously you can't set up a single node ( or maybe you do ) for a single burn and surely after say 5 burns the planet your orbiting will be a day or 6 ahead in it's orbit so how do you know if you're still on target?

Answer:  It simply doesn't matter much, for a few reasons.

First, bear in mind that launch windows are approximate.  Yes, there's one particular moment in time that's the "optimum" launch time... but if you miss it by two or three days in either direction, it really doesn't make much significant difference to your required dV.  So as long as you're not missing your window by weeks, you're fine.  So the fact that you have to do a few orbits doesn't matter much, because each orbit only takes an hour or two even after you've boosted your Ap with periapsis kicking.

"But wait," I hear you cry.  "What do you mean, 'only takes an hour or two'?  If I boost my Ap way high, it could take many days for each orbit!"

To which the answer is:  "Right.  So don't do that." 

Bear in mind that the amount of Ap height you gain per dV goes way up as you approach escape velocity.  For example, from LKO, it takes about 850 m/s to raise your Ap to the Mun's orbit ... but a mere 100 m/s extra will escape Kerbin entirely.   Going to Minmus, at 47K km, only takes about 50 m/s more than going to the Mun at 12K km.

The absolute ceiling on how much dV you can "stash" by periapsis kicking, before doing your final escape burn, is around 950 m/s.  However, you can do most of that, like 700 or 800 m/s, without raising your Ap very high.  800 m/s, which is still sub-Munar, gives you 84% of the maximum possible dV "storage", but your Ap is still down at around 3700 km altitude.

So, what this means is that you simply don't raise your Ap way high.  Raising your Ap higher than 3000 or 4000 km does screw up your departure time by a lot (because the really high Ap means a really long orbit time), but doesn't bring you much dV benefit (because at most you can only eke out another 100 to 150 m/s or so).

So, all you do is set yourself an upper limit of how high you're willing to take your Ap-- say, 3000 or 4000 km-- and that means that your total time spent periapsis kicking won't be all that long.  A few days, at most, which is not enough to screw up an interplanetary launch window. 

[NewtSoup]

Ok, thanks  it's going to take me a little while to get through all this.

 

[NewtSoup]

That's fantastic, thanks. I was suffering from enormous overthink.  Windows are wide,  Oberth effect is advantageous only in the beginning but that's still the bulk of your burn time and you can just keep on making new maneuver nodes.

One of the things that had confused me is that Matt Lowne seemed to be using the "previous orbit"  and "next orbit" tools to create multiple nodes before even starting a burn, but I shall now forget about that completely.  The way you describe it @Snark is much clearer.  Thank you   

I shall try it later this evening in sandbox with a single ion engine craft and compare it to a straight "burn".  After all the best way to learn is by doing,  now I have a better understanding of what to do, I shall do.

 

[Snark]

12 minutes ago, NewtSoup said:

I shall try it later this evening in sandbox with a single ion engine craft and compare it to a straight "burn".  After all the best way to learn is by doing,  now I have a better understanding of what to do, I shall do.

Also, you do know about physics warp, yes?  So you can do your ion burn at 4x normal speed instead of 1x.  Just hold down ALT when you press the "." key.

Apologies if you've known about it since forever and it seems obvious... I only mention it here just in case you haven't been aware.  This game has so many nooks and crannies that it's possible for players to miss some of the less-obvious features even after playing the game for quite a while.  I'd been playing KSP for months, myself, before I found out that physics-warp-in-space was a thing. 

[NewtSoup]

16 minutes ago, Snark said:

Also, you do know about physics warp, yes?  So you can do your ion burn at 4x normal speed instead of 1x.  Just hold down ALT when you press the "." key.

I had no idea.  I've heard physics warp mentioned but I thought it was just doing physics automatically at lower time warps.  Like in atmosphere - I know that's always physics warp.  I think on Linux right shift replaces alt?

[Snark]

1 hour ago, NewtSoup said:

I had no idea.  I've heard physics warp mentioned but I thought it was just doing physics automatically at lower time warps.  Like in atmosphere - I know that's always physics warp.  I think on Linux right shift replaces alt?

Not sure what the story is on Linux key mappings, perhaps someone else could help out there.  What's the key you usually press for all of those KSP features that require holding down the ALT key on a Windows box?  (e.g. toggle thermal overlay, toggle aero display, toggle debug console, vehicle editor part clone, vehicle editor only-attach-to-connector-node, that sort of thing)

Yeah, so, anyway, timewarp.  There are two types of timewarp:  regular, and physics.

Regular timewarp is the thing that happens when you're either landed somewhere, or else you're in "flight" in a vacuum.  The warp where it goes 5x, 10x, 50x, 100x, 1000x, 10000x.  That warp has everything running "on rails" (i.e. the game's not doing any physics calculations), meaning you can't have your engine on, for example, and any craft rotation gets zeroed out, etc.

Physics warp is when the game keeps doing full physics simulation, but at a faster rate.  This is what happens if you're flying in atmosphere, for example.  It goes 2x, 3x, 4x (the max is 4x).  It can be risky to boost faster than 2x, since you can get weird kraken-like glitches (oscillations, wobbly craft, etc.) depending on how much force is involved.

Normally in KSP, the "." key does a physics warp by default if you're flying in atmosphere, but regular warp by default if you're landed or in vacuum.  However, if you're flying in vacuum, you can force it to do a physics warp instead by holding the ALT key (or whatever its Linux equivalent is) while pressing ".".

The advantage here is that you can do that while running your engine, so that you don't have to sit around waiting so long while you're doing a burn.  Use with caution.  If you have a high-TWR craft (e.g. one with a conventional rocket engine on it), there's a chance it could go all noodly/wobbly and tear itself apart.  In my experience, going to 2x is generally safe, but 3x or higher is potentially risky.

However... if you have a really low-TWR craft (e.g. an ion craft), then the forces are so gentle that they're not a problem even at 4x physics warp.  So for cases like that, 4x warp works just fine.  It means you can get through that 12-minute burn in only three minutes. Still kinda long, but it sure saves a lot of tea-sipping and thumb-twiddling while waiting for the burn to complete. 

 

[NewtSoup]

oh yea.  I had a craft just a few days ago turn itself into a circle on change from atmosphere to orbit under high thrust and insufficient struts.  It was kind of bent when it stabilised ( I was amazed it didn't break apart ).  A short burst of warp caused it all to line back up neatly though and I continued on my cautiously merry way.

[Kryxal]

Another way to help with things is to use a planner sat ... plot the node there, plan it so you get to the right point at the right time, then set your burn to match the outbound trajectory.  You'll find that it's a VERY close match at the other end, you might need a few m/s correction burn.  This really helps when there's a significant normal/antinormal component.