Landing next to something

[Elbereth]

I'm trying to build large colonies in different planets and moons, can anyone give me some good advice how to land several different parts from multiple flights into a small area, (preferrably within 50m from each other) without using wheels or flying around too much over the planets surface?

Since the gravity well & atmosphere have quite an impact, I'm starting the first one in Mün

[Azazael]

Hi Elbereth

I use Mechjeb 2.08 for landing. I have 8 x HOME landers all in a cluster within 100m of each other on Duna.

Mechjeb dose a good job of accurately landing as long as parachutes are deployed late. Seems that parachutes deployed early lead to 100-300 metre landing errors.

But as you are going for Mun (no parachute required) try a HOME lander with Mechjeb attached which should be good for <100m landing errors.

[rryy]

Starting from an equatorial, circular, prograde orbit, first, you want to see the latitude of your target, for example 5º south. When you are 90º before passing over your target, you want to burn south so that your inclination is 5º (You can use the Kerbal Engineer Redux or MechJeb to figure out your inclination, or target something already in equatorial orbit). You also want to perform your deorbit burn so that your trajectory lands slightly east of your target. As you are about to pass over your target, begin your burn to kill are surface relative velocity. Then, do your standard landing procedure.

[Sᴄɪɴᴛɪʟʟᴀᴛᴏʀ]

When you are landing on a planet or moon with an atmosphere make sure you compensate for drag by putting your theoretical impact site a little ahead of your intended landing zone.

[Dunbaratu]

I'm fairly sure I'm doing it really inefficiently, but whenever I need a precise landing I prefer to line up a path that would well overshoot the target, but then burn in such a way that I'm not entirely pointed retrograde. I point about 10-20 degrees more horizontally than the retrograde marker, such that I'll end up killing ALL horizontal velocity while still above the target, such that I'm then falling straight down vertically for the last 5-6 kilometers. It's just a lot easier to aim that way. I watch the retrograde marker on the navball and when I see it drift to being exactly centered on the dot on the top of the ball, I stop, let it fall a few seconds, aim straight vertically, and finish the burn that way. This allows me to use a camera view that looks straight down so I can see my target site as I approach.

The other trick I've learned is that if the bottom of my lander was connected to the rest of the rocket using a docking port rather than a sepratron, this gives me a docking port I can rightclick on on the retrograde side of my lander and say "Control from here". Doing that makes everything SO much easier because it means that it mirrors the WASDQE rotational directions, so they behave intuitively in the retrograde direction. When you use the chase camera in a retro-rocket landing, and look at your craft backward toward the ground, if you've selected the docking port on the bottom of the craft as your "control from here", then left will be left, and right will be right, and up will be up and down will be down instead of the brain-hurting inversion you usually get. I mean, having to navigate entirely by navball and not look at the screen lest the backward controls screw you up works fine when you aren't trying to aim your landing specifically, but when you need to watch the screen to correct where you're headed, it's nice to not have the controls mirrored.

One important warning about flying using the bottom docking port as your "control from here" point though: It also swaps the retrograde and prograde markers on the navball! Don't forget that! The prograde mark now shows where the rear of the craft is pointed, so you slow down by burning the engine while pointed PROgrade when it's like this. That screwed me up the first time I tried this and I crashed in the confusion.

Also, if you do this trick and successfully land, don't forget to move control back to the main command pod or probe core again after you touch down, else your ground driving controls (if it's a rover) will really be messed up.

Edited June 7, 2013 by Steven Mading

[Dunbaratu]

Another trick I've done is to send a stationary lander (with legs) with docking ports set up to link up with the docking ports on rovers, designed to line up perfectly with it. This allows rovers to approach the landed lander, dock with the lander on its front and back sides, then you raise the landing gear of the rover such that it's now being supported by the rovers instead of by its own legs. This turns your lander into an unwieldy rover itself, able to move a bit, although it is a bit tippy so be careful and go slow. I usually don't go faster than 3 or 4 m/s with this cumbersome rig arrangement because I'm so afraid of tipping over. After moving the lander, you can then re-extend the landing gear, undock the rovers, and the lander becomes a stationary object again. In this way as long as I land within a kilometer or two of the base, I can have the base send out the rovers to attach to the lander and haul it in. If you're interested I can post screenshots of this arrangement.