How to avoid... well... this.

[Daid]

After the first moon discovered that around an altitude around 1700 there is the moon surface. And moon surface and a 500m/s rocket don't mix.

My 2nd moon lander, well, that's the purple dot at 14.3 km. Or the pod that is left of the moonlander is at 14.3 from the image.

What you see is my 3th moon lander, which I descended very slowly, and at 3m/s it hit the moon, just to bounce off again, gain a lot of horizontal movement, tried to land again and toppled over.

I could not find a way to cancel out the last bit of horizontal movement. Is there some trick to this that I missed?

[duckunlimited2]

No, there's no trick to it. It sounds like you did everything right (descend very slowly, cancel out any horizontal movement, etc). I think the only thing left to do is practice. It will work eventually. You can also help your chances of landing right side up by modifying your design. A wider base and less tall lander can help from tipping over.

[Atimed]

The last few meters you have to tilt the craft little to retro-grade and return it straight back to 0°.. the slower the go the more easily it is to overshot.. requires some practice..

if u can also put a RCS on the ship and use RCS thrusters to negate the horizontal movement..

[duckunlimited2]

I would turn off the rcs upon touching the surface of the mun though. Sometimes, especially when you hit a slope and bounce, the rcs trying to correct the tilt will unintentionally create horizontal movement.

[Volt]

Tripod landing gear are generally a bad idea because the gap between each leg where the craft is free to tip over if it is going that way when it lands is higher. To a certain extent, having more legs gives more stability AND redundancy, but if you have a higher centre of gravity there's not much effect. I always use 4 landing legs, even though I use MJ for my landings.

[Daid]

if u can also put a RCS on the ship and use RCS thrusters to negate the horizontal movement..

Maybe I used the wrong RCS type, but all that RCS gave me was faster rotation controls. Is there some trick on using RCS for horizontal movement cancelling?

Yes, I also noticed the tripods where a bad idea. I've upgraded them to 6 legs for the next ship (which exploded because of the weak decouplers)

[Geredis]

IJKL keys set you up for lateral movements I do believe. They mirror (more or less) the traditional WASD controls for a walking Kerbal. You can also use H or N (again if I remember correctly) to help adjust your descent rate if your craft is small enough/your engines are too powerful/you want/need to save all your liquid fuel for the return flight.

[maltesh]

I almost invariably go with tripod landing gear, but I also tend to design for as wide a footprint as I can, for that higher stability. This was my most recent nearly-stock Mun lander stage:

I suspect I could have landed it nearly anywhere on the Mun without worrying about local slope.

In retrospect, I should have brought more fuel, as the half-tanks turned out to not be enough for even my standard maximum-efficiency return, and I had to go to the RCS to bring them back.

Regardless, I would say test-deploy your lander on the pad without a lifter, and mount those legs as high as you can without risking engine damage. You can mount landing legs on the stock radial engines; this will also help to spread the wide footprint.

Could have putthe legs a bit higher on the body, though.

Edited August 7, 2012 by maltesh

[pebble_garden]

I'd use at least four legs. Tripod gear puts the feet too far apart, and too close to your centerline, making you prone to tipping over.

Cancelling lateral translation just before landing takes some doing, but I've found a technique that works for me.

Try to zero out your lateral speed before you get to 1-2km above the munar surface. You'll still have some, but it won't become apparent until your sink rate gets lower.

Close to hover speed, any remaining horizontal speed will have an alarmingly pronounced effect on your velocity vector. Additionally, it's hard to perfectly center your nose on the zenith dot of the ball, so simple hover-thrusting is going to introduce some lateral component too. The trick is in knowing how to deal with it.

So here you are: slowing to a touchdown with your shadow in view, and your VV (velocity vector) is drifting away from the zenith dot. How do you control it?

I don't really have a name for it, but I will dub it the "Finger/Marble Maneuver". Think of that VV as a marble. Think of which direction you'd have to reach your finger to pull that marble back onto the center of the ball. Mentally draw a line out from the center through that marble, and tilt your thrust vector very gently in that direction. It's not necessary to tilt all the way onto the VV marker, just a few degrees.

Watch as the VV slowly begins to return to the center, as the new thrust vector slowly eats away at it. If it begins to wiggle around to another heading, adjust the angle of your thrust vector. Use small motions and don't overcorrect.

If you don't think you can get the VV centered by touchdown, reorient your ship to pure vertical and thrust enough to climb a bit. This will give you more time to cancel out your lateral drift with this method. But wait until you're descending again to resume finger/marble-ing, as everything would be reversed while climbing.

I found having an ASAS locked while my ship is pure vertical to be very handy. While it is still engaged, I can tilt my thrust vector a couple of degrees, but when I release the rotation key the ship returns to vertical, like a spring. This makes the finger/marble method a lot easier.

Hope that helps. I'm planning to make a tutorial video on this very thing, but I can't resist a cry for help.

PS Also realize that even a tiny amount of lateral speed is going to appear huge when you are in a perfect hover. Don't let that freak you out, it's just math going nonlinear. More important is how much slip you have at around 1-4m/sec.

If you want to see an example of this method in practice, I refer you to this point in my mun landing video, at 19m44s: http://youtu.be/5wlhz7s4J78?t=19m44s I do a big finger/marble maneuver at 20m10s to assume pure vertical flight, and a bunch of wiggling just before touchdown to cancel the last of my lateral drift.

Edit: I should mention I generally don't use RCS for landing, though it could be very handy. My method uses only vectored main thrust, throttled way back.

Edited August 6, 2012 by pebble_garden

[Zephram Kerman]

I agree about adding a fourth leg. That little bit of redundancy comes in handy for me quite often. Even Apollo's Eagle lander had four legs.

Building the lander wide, with stuff on the sides, like Maltesh's footprint example, is the best way to prevent it from tipping over.

[swiftgates24]

Yeah I have a bad experience with RCS units when I landed on the Mun for the first time... I put TONS of them on and when I bounced a little it tried to fix it. I went into an uncontrollable spin D:

[Vanamonde]

In addition to the suggestions above, I like to shutdown my main engine an instant before touching down. I hit slightly harder, but have no thrust adding to that %^(-ing shipkiller bounce. I've lost more ships to that than just about any other cause in KSP. If your ship is tottering, you can hit N to have the RCS thrust straight down, jamming your ship to the surface. As for lateral RCS control, have you tried mounting the "linear" RCS thrusters on the side of your ship? They're stronger than the cluster-type thrusters. I like to put several of them on the underside of my lander so that I can hit H for fine control of descent speed and last-second braking: the more gently you hit, the less you bounce.

[Tigga]

I wouldn't put more than 4 legs on, it looks silly, but 4 is ok.

If you are using ASAS it can cause you issues, especially combined with RCS/thrusters active when you touchdown. I don't find it makes me unstable as such, but if I mess up the landing I find I can abort a lot more easily with it disabled. I tend to just land on the nav-ball using mostly upward thrust (RCS fired linearly, and main engine) rather than using RCS to cancel out sideways motion (tends to roll me annoyingly).

[KevinTMC]

Speaking of bounce...is it just me, or is bouncing more pronounced in 0.16?

I've crash-landed (see sig below) craft that incorporated virtually-indestructible parts (e.g. landing gear, carts), and seen those parts fly a kilometer or more up and away from the crash site.

[Trekkin]

I've done eight legs before when I was using four engine pods. It's handy as a way of making the landing footprint as big as possible, especially for Minmus with its hugely sloped terrain.

It also helps to come in the last meter or so with RCS and ASAS on, since the bounce will be counteracted by the ASAS generally. Even if you do bounce back up, at least it's with the legs down.

[Markus Reese]

Keep your deceleration pointing countered to direction of trouble and finesse your throttle finely to reduce speed even more. If you have any horizontal velocity, it can cause the legs to spring and flippy you over.

[Markus Reese]

Speaking of bounce...is it just me, or is bouncing more pronounced in 0.16?

I've crash-landed (see sig below) craft that incorporated virtually-indestructible parts (e.g. landing gear, carts), and seen those parts fly a kilometer or more up and away from the crash site.

Oh, to answer this, I believe it has to do with changes of the internal collision of parts to help reduce lag. That could be it with the big parts, instead of locking to the model, they sorta bend and flex. Try adding some structural beams to rigid up the structure.

[wired2thenet]

Ditto'd on the fourth leg. Also, strap a MechJeb with Duct Tape on that lander, and when you go for a landing, let MechJeb take the helm for you. Watch what it does and when, then analyze why. Do it two or three times. You'll see the deorbit burns, how it does it and when, and also the final moments before touchdown. First few times I watched it, I saw MJ do some crazy moves that would make Michael Jackson proud, but it touched down safely. Once I saw the results, it took me only two or three tries to duplicate it. Haven't looked back since, and now only use MJ to show me the orbital/surface information (very little automation at this point, although I admit I do it on a few things if it's something I can do in my sleep). Only landings I don't make are the ones due to an errant key press or just "zoning out" at a critical moment.

[Daid]

I'm not using MechJeb, as it feels like cheating myself. I want to figure out myself.

I managed to put my lander upright on the moon, with a silly trick. If you have a low profile lander (with the smaller large fuel tank), and you put 6 legs on it, then you can put yourself upright after tumbling over. Just close your legs after tumbling, and then open them up again. It's a bit of a cheat, but it worked I did break the ASAS from the top of my ship, and the parashute, so no safe return for those kerbals.

[Romfarer]

If you had LazorSystem with Tractor Beams installed, you could use those to flip your vessel up again.

Edited August 7, 2012 by Romfarer

[KevinTMC]

If you had LazorSystem with Tractor Beams installed, you could use those to flip your vessel up again.

Ooh! I didn't know it did that. Thanks for the tip.

Looks like it's past time for me to actually figure out what the LazorSystem does, instead of just looking quizzically at the new parts I installed a week ago...

[kuulness]

You could retract the legs, spin the ship to the side with a leg on it, the try to bounce up useing the legs.