Inclined transfer windows

[Nich]

I am trying to figure out how to do this efficiently.  Idealy I would want to leave Kerbin at the AN/DN and then it would be a straight retro/prograde burn.  However these windows are quire rare as you have to be in the right spot in the orbit and your target has to be in the right spot relative to Kerbin.  So as an alternative you can do a hohmann transfer when the orbits are aligned with a large inclination change or you can do a larger non hohmann burn and intercept at AN/DN with a larger capture burn.  Which is more efficient they both kind of seem like a wash to me.  I also experimented with burning to the purple triangles out of kerbin to push AN/DN so it was more of a hohmann transfer but this required a large amount of dv and I think this would only be worth it if I launched into this orbit straight away.  I find this rather complicated because not only do I have to launch at the right time but at the right angle.

[Snark]

Well, as long as you're not in a hurry and don't mind waiting an extra orbit or two to get to your destination, you can do it like this:

• Launch when Kerbin is at the AN/DN.  Do enough of a burn to move your solar Ap (or Pe, as applicable), to where it precisely intersects the target's orbit.  That will be the opposite AN/DN.  Don't worry about the target planet's position.
• Coast through half an orbit until you arrive ate the opposite AN/DN, i.e. the place where you're at Ap (or Pe) and precisely on the target planet's orbit.
• Now do enough of a burn to adjust your orbit timing so that on the next orbit, you'll arrive at the target planet precisely where you are now.
• If you're unlucky and the target planet's in a really awkward position so that you can't quite get it to line up in one orbit without doing a stupidly large burn, then you can use the dummy-node trick to plan your intercept two orbits from now (or three, or whatever).

Takes longer, but saves a crapload of dV. 

Edited March 9, 2016 by Snark

[Nich]

Ha ha good thinking but unfortunately I am playing with life support so a couple extra years kills on the required supplies.

[Geschosskopf]

2 hours ago, Nich said:

I am trying to figure out how to do this efficiently.  Idealy I would want to leave Kerbin at the AN/DN and then it would be a straight retro/prograde burn.  However these windows are quire rare as you have to be in the right spot in the orbit and your target has to be in the right spot relative to Kerbin.  So as an alternative you can do a hohmann transfer when the orbits are aligned with a large inclination change or you can do a larger non hohmann burn and intercept at AN/DN with a larger capture burn.  Which is more efficient they both kind of seem like a wash to me.  I also experimented with burning to the purple triangles out of kerbin to push AN/DN so it was more of a hohmann transfer but this required a large amount of dv and I think this would only be worth it if I launched into this orbit straight away.  I find this rather complicated because not only do I have to launch at the right time but at the right angle.

Where are you trying to go?  Only Moho, Dres, and Eeloo are inclined enough that it really makes a real difference.  I'm not counting Minmus because there you just leave shortly before its AN or DN with LKO so you arrive at its DN or AN as the case may be, adjusting your Ap to get there at the same time as Minmus.  Hanging out for a few days beyond Minmus' orbit won't bust your life support.

Now, as to the other planets with inclination issues, Moho's a special case with its own set of rules.  There are many threads on fancy tricks for getting to Moho on the cheap which explain it all.

So that leaves Dres and Eeloo.  The way to get to them non-Hohmann is to do a ballistic transfer.  Here, you do the necessary plane-change as part of the transfer burn, to avoid doing a larger plane-change burn en route that the standard Hohmann requires.  To figure out how to do this, there are 2 main options:

1.  MechJeb
MJ's Maneuver Planner ap has an option called "Advanced Transfer to Another Planet".  This brings up a porkchop plot.  Look for the darkest blue areas (zooming in as necessary) as these represent the lowest dV for the trip.  Click on the desired spot on the plot and select "Create Node".  MJ will then create a maneuver node for a ballistic transfer.  This node will be at the proper ejection angle, at the proper point in the future, and will have a bit of normal up/down mixed in with the prograde to send you directly to the target.

2.  Doing it Manually
Use alexmoon's transfer window calculator website  This shows you the same porkchop plot as MJ and tells you how much dV you need in the prograde and normal +/- directions, and what day to leave on.  Then you have to create the node yourself at the proper time, eyeballing the ejection angle, and manually pulling its handles to give it the proper amount of dV in the required directions.  

Even if you use MJ, you might want to run the website anyway while still in the mission planning stage.  That way you'll know the travel time in advance so can pack enough life support for the trip.

Edited March 9, 2016 by Geschosskopf

[Snark]

1 hour ago, Nich said:

Ha ha good thinking but unfortunately I am playing with life support so a couple extra years kills on the required supplies.

Ah yes, that would kinda put a damper on that strategy.

Perhaps one way to have your freeze-dried cake and eat it, too:  Send the heavy stuff (additional supplies, living quarters, mining gear, whatever) unmanned on a cheap but slow orbit, and give it a head start.  Then later on launch your kerbals in a small, light ship that's basically just them, their supplies for the outbound trip, and nothing else.  Because it's small and light, you can build it with a lot of dV and don't have to worry about taking the most-efficient trajectory.  Time it so that they arrive at around the same time (or anytime after) the slow heavy stuff arrives.

Edited March 9, 2016 by Snark

[Nich]

So the specific transfer I am looking at is Eve.  The inclination is small but I would like to learn to do it correctly.  

Do porkchop plots assume an equatorial orbit?

How much dv can be saved by launching into the correct inclined orbit?  I am guessing it is easiest to do a burn to the edge of the SOI 15-20 days before the transfer.  adjust the inclination of the orbit and then burn prograde as you pass PE.

[HebaruSan]

5 minutes ago, Nich said:

I am guessing it is easiest to do a burn to the edge of the SOI 15-20 days before the transfer.  adjust the inclination of the orbit and then burn prograde as you pass PE.

I've done this (I believe for Jool, and again for intra-Joolian transfers), and while it got the craft to its destination and made me feel clever, I did not run the numbers.

5 minutes ago, Nich said:

How much dv can be saved by launching into the correct inclined orbit?

It should be possible to check this without too much hassle by plotting each route with maneuver nodes, then using PreciseNode to get the overall trip Δv. (Stock only shows the cost of the first node, right?)

[Geschosskopf]

1 hour ago, Nich said:

So the specific transfer I am looking at is Eve.  The inclination is small but I would like to learn to do it correctly.

Well, the instructions for using the website are printed right on it, except for 1 thing:  Once you hit the "plot it" button, you get a table of numbers to the right of the plot.  Look on this table for the "Ejection dV".  This is the total amount of dV for the transfer burn along the direction you'll be burning, which of course is inclined to Kerbin's equator, and is not directly useful when creating nodes.  So hit the blue button next to the "Ejection dV".  This brings up a pop-up that has the total transfer burn resolved into its prograde and normal +/- components.  It's those components that you actually use when creating the node.

So, to put this into practice, it goes like this

1. Warp ahead until you're just barely less than 1 orbit away from your departure time (Kerbal Alarm Clock is your friend)..
2. Create a maneuver node at the ejection angle shown by the website.  You'll have to eyeball this unless you use a mod like Protractor or Launch Window Planner.
3. Pull on the node's prograde handle until the amount of dV shown for the burn is the same as the prograde component you got from the website.
4. Pull on the normal up or down (as told by the website) until the total dV shown for the burn is the same as the total Ejection dV from the website.  Compare the predicted trajectory to where you want to go.  Slide the node VERY SLIGHTLY forward or backwards along your orbit, or slightly tweak the burn components, as needed to get the predicted trajectory to hit the target with a Pe approximately where you want it.  This is extremely difficult to do with just the stock interface so you might want to use Precise Node or MJ's Maneuver Node Editor for this fine-tuning.
5. Do the burn.

So to really do this manually without pulling your hair out, you need several instrumentation mods to make the job of node creation easier, even possible.  Or you could just use MJ to do it all for you in 1 go.

There are a couple of caveats either way you go, however.  First, due to floating point errors and the inevitable speedbump when leaving Kerbin's SOI, you will still need to make a small tweak burn en route to get the encounter you set up to start with.

Second, regardless of their inclination, all planets in the game have zero axial tilt.  This means that because you're not hitting the target at AN/DN, you will arrive with some inclination to the target's equator.  Therefore, if you want to be in an equatorial orbit at the target, you must do an inclination burn once there.  These are cheaper the further away from the planet you do them.  Thus, don't do the capture burn all at once.  Stop soon after you've got a closed but highly elliptical orbit just barely within the target's SOI.  This will put your AN or DN with the equator as far from the target as possible.  Coast out there, fix your inclination, then come back and finish your capture burn to achieve the desired orbital altitude.

1 hour ago, Nich said:

Do porkchop plots assume an equatorial orbit?

The website does indeed assume you're in an equatorial parking orbit before the transfer burn.  MJ does not assume this and works regardless of your inclination.  HOWEVER, both assume that your parking orbit is pretty much circular.

1 hour ago, Nich said:

How much dv can be saved by launching into the correct inclined orbit?  I am guessing it is easiest to do a burn to the edge of the SOI 15-20 days before the transfer.  adjust the inclination of the orbit and then burn prograde as you pass PE.

Doing a ballistic transfer saves a little but not a huge amount.  The most savings are for Eeloo or Dres, where you can save a few hundred.  For Eve, Duna, or Jool, it's rather less (a few dozen).  Again, Moho is NOT included in this because you do neither a Hohmann nor a ballistic transfer to get there, but either a bi-elliptic or a gravity brake off Eve instead.  And that's a whole 'nother story.

Note that the dV savings comes in 2 flavors:  The amount needed for the plane change needed to hit the target in the 1st place, and the amount needed for any plane change at the target itself to get into an equatorial orbit there.  The longer you wait to do the en route plane change, the steeper the angle of the plane change, so the more it costs.  Then when you hit the target, you have a greater inclination so the more it costs to get into an equatorial orbit there.  So the ballistic transfer spends less on the en route part by doing it as far from the target as possible, and thus creates the least inclination upon arrival at the target.

The best place to do a transfer burn is at the lowest possible altitude at your departure planet.  The worst place is out near the edge of its SOI.  This is for 2 reaons:

#1:  Oberth Effect
The faster a rocket is already moving, the more efficiently its engines burn fuel.  IOW, it takes fewer units of fuel to create a given amount of dV.  The closer you are to the departure planet, the faster you're going, so the less fuel you burn to generate the dV needed to go somewhere else.  The higher your parking orbit, the slower you're going, so the more fuel you burn for a given amount of dV.

#2:  dV Required
You have to achieve escape velocity to leave Kerbin the 1st place, much less the extra velocity needed to get to the target.  The higher your initial velocity, the smaller the difference between it and your target velocity, so the less dV you need to generate during the transfer burn.  Your initial velocity is highest when you're as close as possible to the departure planet and lower the higher your orbit.

So, the lower your parking orbit, the less dV you need to generate (thanks to higher initial velocity) and the less fuel you burn creating that dV (thanks to Oberth, which is a function of higher velocity).  The higher you are, the more dV you need and the more fuel you burn creating it.

------------------

The thing is, with a ballistic transfer (to anything EXCEPT Moho), the amount of normal +/- dV needed is so tiny that you can do it very close to your departure planet.  Remember, it's not how close you are to the departure planet that matters, it's how close you are to the target.  And your departure planet is as far from the target as you'll ever be, so doing it there is a trivial expense.

 

 

 

[Nich]

Thanks @Geschosskopf that is pretty much what I figured.  I see when I click on the blue Info button it even gave me the heading I need to launch into.  So launching off the pad at the ejection angle with the heading from the info should give the lowest dv transfer (balistic) however this will only save a little dv.  In my case it is a little over 200 dv for Eve (2763 dv total)

[Geschosskopf]

1 hour ago, Nich said:

Thanks @Geschosskopf that is pretty much what I figured.  I see when I click on the blue Info button it even gave me the heading I need to launch into.  So launching off the pad at the ejection angle with the heading from the info should give the lowest dv transfer (balistic) however this will only save a little dv.  In my case it is a little over 200 dv for Eve (2763 dv total)

Hmm, I've never just gone straight from the pad to an interplanetary transfer.  I always go into a parking orbit first because that's what all the algorithms (website, MJ, or any other source) are based on.  IOW, I make a distinction between "launch window" and "transfer window".  A transfer window to me is when the planets are aligned so that the transfer burn requires the least dV, which happens according to the music of the spheres regardless of whether you're on the ground or in a parking orbit.  I use "launch window" strictly for the case of when to leave the ground to go directly from the ground to some target orbiting that same planet.

Anyway, I don't think the "heading" in the pop-up on the website is the direction to launch at.  As mentioned, the website assumes you're already in a circular, equatorial parking orbit, which means it's assuming you launch at 90^.  If you just hit "Plot It" leaving all the default values in place for the 1st Duna window, the heading will be 78^, which will put you in a Kerbin orbit with 12^ inclination.  This is absurd given that Duna has only a 0.6^ inclination.  Rather, I think that "heading" in the pop-up is a sanity check that you created your node correctly.  I think that when you create the node at the proper ejection angle, its marker on the navball should be at 78^.  I THINK.  But it's definitely not your launch azimuth.

For instance, the dV maps (which do not lie) and long experience indicate that the transfer burn from Kerbin to Eve requires only 1490m/s worst case.  That's if you hit Eve when it's at its highest point above or below the ecliptic.  If you happen to hit Eve at an AN or DN, then it's only 1040, barely more than going to Minmus.  Usually, however, the transfer is about 1100-1200m/s.  So a 2700+ transfer burn is way off, and is probably a result of being in an overly inclined Kerbin orbit and having to drag your inclination back down to hit Eve.

[PLAD]

6 hours ago, Nich said:

So the specific transfer I am looking at is Eve.  The inclination is small but I would like to learn to do it correctly.  

Do porkchop plots assume an equatorial orbit?

How much dv can be saved by launching into the correct inclined orbit?  I am guessing it is easiest to do a burn to the edge of the SOI 15-20 days before the transfer.  adjust the inclination of the orbit and then burn prograde as you pass PE.

Alex Mun's porkchop plotter gives the departure thrust required assuming you are starting from an equatorial orbit. This will usually be a higher number than if you are allowed to start from an optimally inclined orbit, but since Kerbin's surface is moving at 175m/s at the equator it will cost more to get from the KSC into a non-equatorial orbit than to launch into an equatorial orbit.  So we come to your question about how much launching into the optimal orbit can save in the departure burn.  Looking at the first Kerbin-Eve window,  Alex Mun's plotter says the lowest-dv transfer (from an equatorial orbit) leaves Kerbin on Y2 D160.8 and arrives at Eve on Y2 D357.4. From an equatorial, 75x75km orbit it requires a 1072m/s ejection burn (Normal of -201.8m/s, prograde of 1052.4). Note that this would be one burn, done at 75km Kerbin altitude. It would be very inefficient to do a 2nd burn near Kerbin's SOI to correct inclination.
    I wrote Flyby Finder to compute departure burns assuming you start in the optimally inclined orbit.  If you search the same window with it it gives the lowest dV transfer (from any inclination orbit) as leaving Kerbin on Y2 D141 and arriving at Eve on Y2 D339 and requiring a 1058m/s ejection burn (all prograde), but you have to start from an orbit that is inclined -28.3 degrees.  I tried it and found that when launching a Kerbal X into these orbits with Mechjeb it costs me 27m/s more to get into the 28-degree 75x75km orbit than an equatorial orbit. So the equatorial is better by a little bit, unless for some reason you need to leave earlier in the window. Note that Mechjeb isn't too accurate in placing you in an exact inclination and LAN orbit, (and I'm far worse trying to do it manually) so the equatorial approach is usually better in practice since that is easier to get into.
This is normally  the case, either way is pretty close to each other in total dV, so the easier equatorial wins.  The exception is when there is no low-inclination solution in a given transfer window (which happens sometimes for the slow-moving Dres, Jool, and Eeloo) or any time you need to arrive at your destination when it is far from Kerbin's orbital plane. Then you can save up to hundreds of m/s by launching into the optimally inclined Kerbin orbit instead of an equatorial Kerbin orbit.
And if you are using RSS, or in the real world, then all of this goes out the window since you launch from a non-equatorial site, and the Earth's equator isn't in the plane of the planet's orbits. Here a properly-timed due-East launch is almost always the way to go.

[HebaruSan]

20 minutes ago, PLAD said:

From an equatorial, 75x75km orbit it requires a 1072m/s ejection burn (Normal of -201.8m/s, prograde of 1052.4). Note that this would be one burn, done at 75km Kerbin altitude. It would be very inefficient to do a 2nd burn near Kerbin's SOI to correct inclination.

20 minutes ago, PLAD said:

If you search the same window with it it gives the lowest dV transfer (from any inclination orbit) as leaving Kerbin on Y2 D141 and arriving at Eve on Y2 D339 and requiring a 1058m/s ejection burn (all prograde), but you have to start from an orbit that is inclined -28.3 degrees.  I tried it and found that when launching a Kerbal X into these orbits with Mechjeb it costs me 27m/s more to get into the 28-degree 75x75km orbit than an equatorial orbit.

Apologies if this is already covered by your analysis and I missed it, but what about Option #3: launch to equatorial, then split the transfer burn in half (or smaller) and change planes in between at apoapsis? I know plane changes <45° aren't made more efficient by raising and lowering apoapsis, but if you're going that way anyway, is it a better idea?

[Nich]

So 2770 dv include powered capture at Eve.  I was working under the assumption that aerocapture at eve is impossible because you basically explode the second you hit the atmosphere at interplanetary speeds from what I have heard.

 

When I said burn at kerbins SOI I meant in a highly elliptical orbit still in the SOI.  This way rather then having to do a -201.8 m/s normal burn you can do a 2 or 3 m/s normal burn at AP then when you come back to PE you can do the 1052.4 prograde burn.  But in reality thanks to pythagorean therom your -201.8 m/s normal burn only cost you 20dv.  In addition this manoeuvre adds significant risk as you now have to account for kerbins rotation of 360 degrees / 426 days = .845 degrees/day.  Over the 20ish days it takes to get to the SOI and back you would have to add/subtract 17ish degrees.  For the 18ish dv this would save you it is hardly worth the risk.  Interesting I set up my plot from a 72 km kerbin orbit to a 92 km Eve orbit.  I got a 1069.8 prograde and -199.8 normal burn at 156 to retrograde with a heading of 100.58 and  Total dv of 1088.  In theory I could launch at 156 retrograde with a heading of 100.58.  Then refuel and burn to Eve entirely prograde saving 17 dv.  However for the complexity this adds in refueling and the fact that I am packing an extra 250 dv margin this is hardly worth it.  

 

In the future I think I am more inclined to take a mid coarse plane change as it looks like on this transfer there is a 1417 dv capture requirement for plain change vs 1667 dv requirement for ballistic.  I dont plan on aero capturing but 250 m/s could be the difference between a 500 m/s aerobrake and a 100 m/s aerobrake

[PLAD]

1 hour ago, HebaruSan said:

Apologies if this is already covered by your analysis and I missed it, but what about Option #3: launch to equatorial, then split the transfer burn in half (or smaller) and change planes in between at apoapsis? I know plane changes <45° aren't made more efficient by raising and lowering apoapsis, but if you're going that way anyway, is it a better idea?

Oops, you're right, I didn't address that method. Excuse me while I figure this out.

1) Get into a 75x75km equatorial orbit. This saves 27m/s over launching into a 28-degree orbit.

2) Get into a 75x 83,000km orbit, which goes right out to the edge of Kerbin's SOI. This will require a boost of 934m/s, and the velocity will be 26m/s at apoapsis. Because Kerbin's SOI does not go to infinity you cannot reduce this number and stay in Kerbin's SOI. Note that you will require several adjustments to get this apoapsis so precise, in practice you will blow all your gains in these maneuvers, or you can settle with a lower apoapsis, but this is an ideal case.

3) At apoapsis change the plane of the orbit by 28.3 degrees. This will require a ~12.5m/s thrust. (Law of sines)

4) at the next periapsis complete the prograde thrust, in this case to get from 934m/s over circular V at 75km to the desired 1058m/s, so another 124m/s and you're on to Eve.

Now you have reduced the 27m/s extra required for the 28-degree solution to 12.5m/s extra. Since the 28-degree departure cost 14m/s less (1058 versus 1072) this means the 28-degree solution is very slightly better, but given that it takes about 250 hours to get out to the SOI and back and that course corrections will cost you more than you save I would stick with the equatorial solution. This should get much more useful for higher inclinations, but  I think it would have to be examined case by case rather than some rule of thumb though.

 

[Plusck]

I have tended to eyeball the inclination from launch, recently. It helps to have at least one craft in LKO to have a nice line along Kerbin's equator

Basically, you select the target planet as target, then rotate the map view around until the target planet's (green) orbit intersects an equatorial LKO orbit. Then you want to launch into orbit, shortly before midday or midnight, at exactly the opposite angle compared to the equator.

If the target orbit is not at AN/DN (so when looking at the equator on Kerbin, you can see both the near and the far sides of the orbit) then your launch angle should be slightly less than the "opposite" of the near, prograde side of the target orbit.  

I'm not too sure of exactly how much fuel you save doing this, but I doubt it is much (maybe 100 m/s or so for Eve, maybe more for Moho).

[Geschosskopf]

On 3/10/2016 at 7:42 PM, Nich said:

So 2770 dv include powered capture at Eve.  I was working under the assumption that aerocapture at eve is impossible because you basically explode the second you hit the atmosphere at interplanetary speeds from what I have heard.

Yeah, it takes 1200-1500m/s to capture into a low Eve orbit by thrust alone, and aerocapture there from an interplanetary trajectory is pretty much suicide.  You can sometimes save a little on the capture dV by aerobraking on subsequent orbits after thrust-capturing on the 1st orbit.  Use thrust to bring you Ap down to about 1/2 to 2/3 of the way down from the SOI edge, then go there and bump  your Pe into the atmosphere.  This still takes a pretty heat-tolerant ship, probably with a fairing.  In which case, the sleekness of the fairing will minimize the speed you lose in aerobraking, so this either takes many passes or just doesn't work at all.  And you'd only save a few hundred m/s so it's usually simpler to just pack that much fuel and burn the whole capture.

Now, with the new inflatable heatshield in 1.1, things hopefully will be different.

On 3/10/2016 at 7:42 PM, Nich said:

When I said burn at kerbins SOI I meant in a highly elliptical orbit still in the SOI.  ... However for the complexity this adds in refueling and the fact that I am packing an extra 250 dv margin this is hardly worth it.  

In the future I think I am more inclined to take a mid coarse plane change as it looks like on this transfer there is a 1417 dv capture requirement for plain change vs 1667 dv requirement for ballistic.  I dont plan on aero capturing but 250 m/s could be the difference between a 500 m/s aerobrake and a 100 m/s aerobrake

You're really over-thinking this.  Except for Dres and Eeloo (again, ignoring Moho), the target's inclination is so slight that you don't need to worry about saving fuel on it with added complexity.  The size of the stock KSP solar system (even if you add the Outer Planets Mod) so small that it really doesn't take much dV to go anywhere you want by brute force and ignorance.  The tiny amount of fuel involved in the plane change necessary to go to Eve, Duna, or Jool will not make or break your rocket design, and certainly isn't worth adding any complexity to the mission to try to save.  Just start in an equatorial orbit, plot a ballistic transfer, and problem solved :)..  And even for Dres and Eeloo, while you might save a bit more, they're so far away that doing the plane change as part of the transfer burn is still an insignificant cost not worth extra complexity.

The only time when getting fancy with a transfer burn is worth doing is when you break one long burn up into shorter segments.  You never want to burn more than about 5 minutes or you lose lots of efficiency to the Cosine Gods and might even push yourself into the atmosphere at Pe.  So, if you have a transfer burn of say 8-10 minutes, you can do the following:

• Get into a parking orbit a couple days in advance of your departure time.
• Create the transfer node at the correct ejection angle.
• Burn about 1/2 the transfer dV as you pass the node, raising your Ap and putting your Pe where the node is, until the remaining burn time is about 5 minutes.  You just have to be careful that you don't end up getting grabbed by Mun while you wait for the departure time.
• Delete the original node and create a new one for the remaining dV.  This is because the original node was assuming you were in a circular orbit.  Now you're in an ellipse so will be moving faster as you go through Pe.  Also, your orbital period is longer so you won't reach the node at the same time you would have before.  This will change your departure time a bit but not enough off the optimal solution to matter..

This is really only practical for original burn times of about 8-10 minutes.  If it's longer than that, then you're pretty much stuck with starting from a higher orbit (where the burn time no more than about 20% of your orbital period).  That means accepting a loss of efficiency due to less Oberth and lower initial velocity, but if your TWR is so low that your burn time is > 10 minutes, it's usually because you have all the dV in the known universe so don't care about efficiency anyway.

[Warzouz]

On 9/3/2016 at 5:45 AM, Geschosskopf said:

...

1.  MechJeb

MJ's Maneuver Planner ap has an option called "Advanced Transfer to Another Planet".

...

Beware that MJ calculates a efficient burn to leave origin SOI. It doesn't calculate good slowdown encounter. Only Alexmoon site does.

Maybe, I'm not using MJ correctly, but nearly any time I let it calculate a Moho encounter, I got a very bad one with a non tangent intersect. Nearly every time the MJ solution was much less efficient than a equatorial burn + plane change (If I had done the slow down at target).

I wouldn't recommend using MJ for more than a Fly-by.

[Geschosskopf]

33 minutes ago, Warzouz said:

Beware that MJ calculates a efficient burn to leave origin SOI. It doesn't calculate good slowdown encounter. Only Alexmoon site does.

I believe you're correct.  However, if you build your ship according to the worst-case scenario shown on the dV map, then this isn't a concern because you already know you have sufficient fuel for the trip.  I just use MJ to create the node for me because doing it manually is such a pain with the stock interface, or even with helpers like Precise Node and such.  The whole map view and maneuver node system was designed for going to Mun and just doesn't work well at interplanetary distances.  I certainly hope the UI revamp in 1.1 addresses this.

33 minutes ago, Warzouz said:

Maybe, I'm not using MJ correctly, but nearly any time I let it calculate a Moho encounter, I got a very bad one with a non tangent intersect. Nearly every time the MJ solution was much less efficient than a equatorial burn + plane change (If I had done the slow down at target).

Moho is a thing unto itself.  If you want to get there without massive amounts of dV, do NOT go there using a conventional transfer (ballistic or Hohmann) and do NOT leave on a conventional transfer window.  While you can do things that way, and the transfer burn is reasonable, you will usually find that the cost of the capture burn is like 4000-6000m/s.  If you want to go there on the cheap, do a bi-elliptic transfer leaving Kerbin at its AN or DN with Moho.  Slightly more expensive but still way cheaper than going there conventionally is to leave Kerbin on an Eve window and do a gravity braking pass off Eve.  With both methods, you will probably have to do 1 or more solar orbits to sync up for a Moho encounter because you leave Kerbin without regard to its position relative to Moho.  Thus, the trip usually takes 1-2 years.  However, only by doing one of these methods will the dV cost approximate the rather low numbers shown on the dV map.

33 minutes ago, Warzouz said:

I wouldn't recommend using MJ for more than a Fly-by.

I totally disagree.  MJ is fully capable of plotting all types of orbital maneuvers, saving you from having to struggle with the inadequate interface.  It can also, if you want, execute the burns more precisely than you can manually.  HOWEVER, that's all it does. It doesn't build the ship and it doesn't offer advice on how to fly it.  That's all on you as the player.  You have to use the dV map to design the ship with sufficient dV and TWR for the intended mission, and you have to know orbital mechanics well enough to know what type of maneuvers you need to perform at various points along the way.

MJ is best thought of as a subordinate crewman or worker.  When building the ship, it's like an engineering intern constantly recalculating dV and TWR as you swap parts around so you know when your design meets requirements.  But you have to establish those requirements ahead of time when designing the mission.  When flying, it's like Mr. Sulu and you're Capt. Kirk.  You have to know what maneuver you need to do and how best to do it before you can order MJ to perform the donkey work of setting it up and/or executing it.  IOW, to be able to use MJ effectively, you have to be capable of doing it all yourself without MJ.

[Warzouz]

2 hours ago, Geschosskopf said:

I totally disagree...

I wasn't clear enough and/or you didn't understand what I wanted the say.

I was speaking about transfers. Sure MJ is a very useful mod. I use it a lot to do basic manoeuvres like circularize. I use it for rendez-vous because I'm bad at it. A use the "Execute node" with around 80% of my own planed nodes. The node editor is not as nice as Precise Node, but it's efficient.

In Beta, I used to go to orbit using the ascent guidance, but since 1.0, I find that flying rocket is much more interesting than before.

So to speak, I use a lot of MJ features but I don't use it to get planetary transfers.

 

[Geschosskopf]

1 minute ago, Warzouz said:

So to speak, I use a lot of MJ features but I don't use it to get planetary transfers.

OK.  But it does work quite well for that, and is a lot more convenient than wrestling with the horrific interface.

[Warzouz]

2 minutes ago, Geschosskopf said:

OK.  But it does work quite well for that, and is a lot more convenient than wrestling with the horrific interface.

True, but I use MJ "Node Editor" and I'm quite happy with it.

[Nich]

Thanks for all the help I have launched my Eve mission if you want to check it out.

 

[Laie]

On 3/9/2016 at 6:06 PM, Nich said:

So the specific transfer I am looking at is Eve.  The inclination is small but I would like to learn to do it correctly.  

Do porkchop plots assume an equatorial orbit?

How much dv can be saved by launching into the correct inclined orbit?  I am guessing it is easiest to do a burn to the edge of the SOI 15-20 days before the transfer.  adjust the inclination of the orbit and then burn prograde as you pass PE.

Alex Moon's porkchop plotter (and the Transfer Window Planner mod that does the same in-game) both assume that you start from, and want to enter into, an equatorial orbit. Mechjeb's Advanced Transfer planner, however, bases it's prediction on your actual orbit.

As has been said a few times already, inclinations in the stock game usually don't matter very much. The way vectors add up, a small inclination change on top of a standard transfer barely affects delta-v. When going to Jool, you may need to do 2000 prograde and 200 normal for a grand total of 2015; if you were to start from the right inclination, you could do a prograde-only burn but would still need 2010.

The best way to save dV with plane changes is to plan you mission so that your AN/DN coincides with a point where you want to do a burn anyway; one can "hide" a sizeable normal component in the insertion burn and shave off a few degrees at basically no cost.

[Nich]

So I did have an additional question.  I got the intercept pretty easily but getting my PE down to 95 km was quite challenging.  I know that time is quite important however I would think that 1/2 hour or even a day would not make much or any difference as the time scale is so small compared to the rotation.  I am very good at tweaking rendezvous (I have actually collided with a couple stations because I was not paying attention) Has anyone seen a tutorial on how to push things around with a ballistic intercept?  Burn angle, Radial, normal and prograde all seamed to have a lot of cross coupling.

[Plusck]

On 09/03/2016 at 6:18 PM, HebaruSan said:

I've done this (I believe for Jool, and again for intra-Joolian transfers), and while it got the craft to its destination and made me feel clever, I did not run the numbers.

It should be possible to check this without too much hassle by plotting each route with maneuver nodes, then using PreciseNode to get the overall trip Δv. (Stock only shows the cost of the first node, right?)

Well, you can mouse over each node to see the dv cost.

I was playing for months, deleting each node to see the value of the next one, before I even realised this...