Tricks & Tips for refining interplanetary maneuver nodes

[brdavis]

I'm a physicist, so I've got a reasonable grasp on how to do an interplanetary transfer; and the maneuver nodes make setting things up pretty easy (yes, I'd like ejection angle readouts too; but I'll deal with it ). But one thing that bugs me is trying to see the effects of adjusting a maneuver node where the proposed trajectory intersects a distant SOI. As an example, I was 4 days out from intersection Kerbin's SOI on a Duna return trajectory and about all that could be seen was that things were changing… I had to fine-tune my entry trajectory by tweaking the maneuver node while mousing back over the Kerbin periapsis marker to figure out if at was getting smaller or larger… and then when I found the minimum in one axis (say, pro/retrograde) I'd have to repeat the searching in a different axis (up/down or radial). It's like a nearly blind optimization search of a five-dimensional phase space (X/Y/Z time of maneuver and delta-v magnitude).

Suggestions? Tips? Tricks?

Up until now I've often used the inefficient method of multiple maneuvers (boost to orbit ==> eject to edge of SOI ==> adjust inclination ==> transfer to Hohmann transfer orbit, etc.). I know that's inefficient, but the frustration of not being able to "see" what effect I have in the (often distant in both time and space) SOI is really a difficulty.

I don't use MechJeb or Engineer (although the Edu mod does give you some of Kerbal Engineer functionality, and I'm testing that), so I'm looking for the best way (not the perfect way) to do this with a mostly stock install.

Oh yeah, RCS makes interplanetary orbital refinements much more enjoyable….

Edited May 21, 2014 by brdavis

[Dave Kerbin]

In stock you can, if you hit it in the right place, click the PE marker to 'sticky' it. However if you are modifying a maneuver node and even briefly break the encounter or lower PE to 0 you'll lose the sticky.

I tried for a long time to do it pure stock but after some frustration at Eeloo I finally installed PreciseNode. Among other things it lets you adjust the maneuver using the keyboard so you can keep the mouse where you want it. It can also optionally show the ejection angle (there is an O button for options that lets you hide/show various features). Finally it has a conic mode feature to let you change how projections are shown. I don't know what I changed it to but I like it more then what it was before (I now have it showing interception paths in the same frame of reference as the body they are intercepting, so for example a Mun encounter isn't a line that bends bit as it encounters the Mun and then goes back, it sharply twists around to show the half orbit, then breaks off to continue elseswhere)

[Starstrider42]

In 0.23.5, you can center the map view on any planet (either by cycling through them with tab, or by left-clicking the planet and selecting "focus") and all trajectories, including maneuver node trajectories, will be redrawn relative to that planet. I don't have a screenshot handy to demonstrate, but it's quite helpful. It's pretty easy to plan a ~1 m/s course correction with the new display.

Best ARM feature ever.

[Yasmy]

ninjaed...

If you are not using 0.23.5, a stock solution is to change the patch mode to zero in settings.cfg:

CONIC_PATCH_DRAW_MODE = 0

Then use Tab to focus on the body you want to refine the maneuver around.

Edited May 20, 2014 by Yasmy

[allmhuran]

Yes, it's a problem when you have to "look through the sun" to keep an eye on both your target periapsis and the node handles, and from a very horizontal viewpont the handles sometimes seem to "overlap" in terms of which one you are clicking.

I suggest downloading "precise node" mod. It pops up a little dialog window when you add a maneuver node, so you can stay zoomed in on your target periapsis and use the buttons in the dialog to adjust delta V.

[Taki117]

In 0.23.5, you can center the map view on any planet (either by cycling through them with tab, or by left-clicking the planet and selecting "focus") and all trajectories, including maneuver node trajectories, will be redrawn relative to that planet. I don't have a screenshot handy to demonstrate, but it's quite helpful. It's pretty easy to plan a ~1 m/s course correction with the new display.

Best ARM feature ever.

You could do this in other versions too, and this is by far the best option.

[Kobymaru]

I use Alexmun's Launch window planner for the approximate data, and then fine-tune the maneuver node using MechJeb's maneuver node editor.

I guess on a stock install it's going to be difficult, because you dont have precise "angle to prograde" values and fine-tuning will be difficult.

I do have a small tip though: starting from .23.5 you can zoom in on the target SOI and fine-tune your burn using your mouse-wheel when you put the mouse pointer on the prograde/radial/normal markers.

I suggest downloading "precise node" mod. It pops up a little dialog window when you add a maneuver node, so you can stay zoomed in on your target periapsis and use the buttons in the dialog to adjust delta V.

I checked it out, but isn't most of the functionality already included in MechJeb's "Maneuver Node Editor"?

[allmhuran]

I couldn't say, I've never tried mechjeb

[Kasuha]

Alexmoons launch planner is definitely the tool to use.

To set up maneuvers in 0.23.5, you can now adjust maneuver parameters in cm/s using mouse wheel while pointing your mouse at maneuver icons (needs some practice, mouse wheel step causes very different value changes depending on how fast you scroll).

But: do NOT trust your maneuvers. Rather than executing the burn to kill the last cm/s of your prepared maneuver, delete the maneuver shortly before you finish the burn and watch what happens with your trajectory in map view. Then, similarly how you were refining your planned trajectory using mouse wheel, you can now refine your real trajectory using RCS burns. This allows you to prepare Laythe intercept while you are still less than 100 km from Kerbin and adjust it to aerobrake right after you leave Kerbin SOI.

Recommended setting: CONIC_PATCH_LIMIT=6 or more

[nli2work]

Mechjeb has maneuver node editor that lets you fine tune in 0.01; 0.1; and 1 increments.

But as Kasuha said, don't trust maneuvers completely. you'll have to make adjustment burns mid course once or twice at least. I've had trajectory change, sometimes quite a lot, just from switching to another vehicle and back, after I made a burn.

[phoenix_ca]

Vagaries of floating-point calculations. Corrections will be necessary.

[Starstrider42]

You could do this in other versions too, and this is by far the best option.

Then how come it's listed in the ARM changelog?

* Celestial Bodies now show a 'Focus View' button on their context menus in Map View.

...

* Celestial bodies you have an encounter patch with will display that patch in local mode when they are focused on the Map View or Tracking Station.

[brdavis]

Thank you all - some of these I already knew, some were new. I've manually changed the conic type and zoomed in on the SOI to see what the result was, but the problem is trying to keep that up and visible and still have access to all the handles on the node itself (which may be quite far away, or as another mentioned "seen through the Sun").

I ended up adding PreciseNode and Kerbal Alarm Clock, both of which *greatly* decrease the frustration. MechJeb would probably be good… but like some others, it's not my style of play. AlexMoon's calculator is great as well (combining that with PreciseNode let me set up a Jool intercept from 100 km Kerbin orbit the very first time… impressive use of both tools!).

[RW-1]

In 0.23.5, you can center the map view on any planet (either by cycling through them with tab, or by left-clicking the planet and selecting "focus") and all trajectories, including maneuver node trajectories, will be redrawn relative to that planet. I don't have a screenshot handy to demonstrate, but it's quite helpful. It's pretty easy to plan a ~1 m/s course correction with the new display.

Best ARM feature ever.

This, though if it doesn't suffice, then precise node is an option. drag it out to hit SOI, focus on planet and adjust, neat to see as well IMHO.

- - - Updated - - -

In 0.23.5, you can center the map view on any planet (either by cycling through them with tab, or by left-clicking the planet and selecting "focus") and all trajectories, including maneuver node trajectories, will be redrawn relative to that planet. I don't have a screenshot handy to demonstrate, but it's quite helpful. It's pretty easy to plan a ~1 m/s course correction with the new display.

Best ARM feature ever.

This, though if it doesn't suffice, then precise node is an option. drag it out to hit SOI, focus on planet and adjust, neat to see as well IMHO.