How do you fight Lander Tipping Disorder (LTD)?

[Vallius]

It's happened to us all at least once. You bring down your carefully designed lander to the surface of a new world. You've only got about 500 m before you make history, when you suddenly realize that the "flat" landing spot is actually a 30 degree slope! You scramble to reorient your craft, try to shift it over a bit to a flatter surface, but it's no use. You hope against hope that your meticulously planned mission will get a lucky break

In retrospect, not the best choice of words...

, and touch down at 5 m/s--horizontal velocity.

If you're really lucky, Jebediah was driving and you can manage to salvage enough of your ship to get back into orbit.

If your fortune was like my latest Mun mission with Munsen Kerman, your crew will survive, but your ship will break to bits.

How do you veterans avoid or mitigate Lander Tipping Disorder (LTD)? I've started designing ridiculously wide landers, and I've seen a good tutorial suggesting making final landing from a 5 km orbit (on the Mun).

Any advice for landing on other bodies?

Edited October 31, 2014 by Vallius
Typos...

[Master Tao]

Wide, low landers with enough fuel to hover just before touchdown:

I believe that was a 40 degree slope. The LV-909 let me play with the throttle very carefully and Jeb made sure to touch down on the high side first. A little horizontal velocity into the slope can help, too. In general, you want your science experiments above the capsule, since they quite a bit less – a Science, Jr. is less than 1/3 the mass of a Mk. 1 Lander Can.

[Alshain]

1. Wider not Taller

2. Low Center of Mass

3. Landing gear mounted as close to the center of mass as possible without bottoming out the engine or other components.

Radial engines help, test the landing gear deployment in the VAB, like the picture above it should be just off the ground. Don't position the landing gear at the very bottom with enough space to drive a rover under it (unless you are deploying a rover under it).

[AlexinTokyo]

In terms of design:

• Wide base - you can use girders or radial decouplers to extend landing legs further outside the body if need be.
  
• Low CoM - put the heavy stuff at the bottom and the light stuff at the top, and don't build too tall.
  
• RCS - properly balanced RCS thrusters so you can impart/cancel horizontal velocity in both planes without needing to rotate away from vertical.
  
• Contingency fuel - pack enough propellant to allow for hover time while choosing a landing site.
  

In terms of piloting:

• Constant altitude landing - gives you more chance to spot a good landing site.
  
• Zero horizontal velocity - reduce the horizontal velocity relative to the surface to zero before landing.
  
• Very low landing speed - on slopes, reduce vertical velocity to as close to zero as possible, this can help prevent the up-slope legs from bouncing and tipping the lander.
  

Also, mods like KER, MJ, VOID, and Trajectories can help by giving information on the slope under the lander, and by enabling more precise choice of landing sites.

[Kerbolnaut]

Design: It definitely helps to make a wide lander. I usually make landers with three radially-decoupled fuel tanks and small engines around the return module. I place the lander legs on the decoupled fuel tanks so that the lander has a wide stance which can safely tolerate shallow slopes. It also helps to keep your lander short so that it has a low center of gravity, and that's easier to do when the spare fuel is kept to the sides rather than vertically.

Planning: It's usually a good idea to establish a low orbit and swing around once or twice to scope out possible landing sites. It's also generally best to attempt landings on the current daytime side so that you'll have good illumination to help you spot slopes and avoid them early on.

Piloting: If you've picked out a landing site that seemed okay, but looks sloped when you get closer to it, then it's best to fire up the engines and move a little bit so you can get a safer landing site. Apollo 11 had to do exactly this when Neil Armstrong noticed that their intended landing site actually had a lot of boulders strewn about, so he was forced to make the landing a short distance away from the original target. As always, make sure to get rid of horizontal velocity before trying to go down to the surface. Have SAS on (and RCS if available) in the last few seconds before you touch down so that rotation induced by the landing can be minimized.

Recovery: If you've done your very best to avoid tipping on the landing but you still find that you're about to tip over, you should immediately go to full throttle to bounce yourself away from the surface. You'll most likely need to correct a tumble, and you'll often have less than 30 seconds to do it. If you don't maintain full control during the bounce, you should kill the throttle as soon as the lander is aimed more down than up, then stabilize your lander, give yourself a short burn upwards to buy some extra time, correct any horizontal velocity you might have picked up and try again. When in doubt, just remember that it's always safer and less expensive to retrieve kerbals from orbit than it is to retrieve them from the surface of a planet or moon. As such, don't worry too much about wasting a little fuel on the landing - it's better to err on the side of caution and get the landing right.

[cantab]

Yeah, hitting the ground with horizontal velocity can tip over even a wide flat lander. So watch the navball. Using RCS to fine-tune horizontal velocity can help too.

[tsotha]

Hmmm, well, that's what's great about landing on Minmus. Sometimes I don't even make a "lander" - I just put a couple SAS modules on my rocket and lay it down wherever I happen to end up. When it's time to leave I use the torque to lever the thing straight up and fire the engines.

The others are right, though. Wide landers with a low center of gravity are the way to go. Make sure you kill all your horizontal velocity, and the steeper the slope the slower you need to touch down. SAS helps too, since it means you can put all the weight on the up-slope legs and gently lower the others.

For a basic Mun lander, instead of having one tall tank I'll split my fuel into four tanks and attach three radially from the center. The legs go on the outside, so the thing ends up looking more like a spider than a tower. Also, you don't need to attach the legs flush against your ship - if you have the clearance you should angle them outward to increase your footprint even further.

[RexKramer]

What I would add to the previously stated is I rarely put landing legs on landers. Almost never, actually. Generally, if you get your velocity to less than 5 m/s, you won't damage engines or whatever is on the bottom. The impact tolerance of parts is listed in the VAB menu.

Otherwise, the above suggestions all work well. Build wide and low landers. Radial mounted tanks help keep your center of mass low. Having adequate SAS helps also- usually the command module's SAS is sufficient, but for heavier landers adding SAS modules doesn't hurt.

For really bumpy terrain which has caused you problems in the past: Don't plan a steep final descent which leaves you with few options to change your landing site. Instead, plan enough fuel to skim the surface maybe 1000m up, at a modest horizontal speed. When you find a flat(ish) landing spot, stop your horizontal motion and drift down. This takes more fuel than a suicide-burn drop, but leaves you more options.

[steve_v]

The same way RL landers do it: carrying enough fuel to land somewhere flat

Using SAS is cheating IMHO.

Edited October 31, 2014 by steve_v

[DrMonte]

Make your lander untippable (if such word exist).

With no intention I made one so wide that a Kerbal was not able to stand near it. Landed at the Mun pole (you know how the land is there).

[KerikBalm]

I find it useful to activate RCS upon touchdown.

Obviously, couple it with lots of RCS ports.

[Leszek]

Step 1: Kerbal Engineer, the surface window has a slope indicator for your touchdown spot.

Step 2: Land slowly.

Step 3: SAS.

[whaaw]

Step 1: Kerbal Engineer, the surface window has a slope indicator for your touchdown spot.

^

|

|

This!

[Brainlord Mesomorph]

I toyed with horizontal landers for a while.

Designed to land on their side. Reoriented the landing struts for the side. And added “vertical ascent thrusters †smaller rockets designed to lift it up off of its side long enough for the main engines to kick in.

The small unmanned one I called “science dog.†And the larger manned version ended up being almost precisely a Space 1999 Eagle. I had to build them in the SPH, to use that kind of symmetry.

Flew these things all around the KSC for quite awhile. They were fun to fly. Very science fictiony. (Wes Anderson would have been proud)

When I sent them on missions though they had problems:

Switching from the vertical to horizontal orientation during landing was difficult. And I realized that I was carrying around six tons of vertical thrusters that I only use for like 3 seconds in a mission.

It’s just wasn’t efficient.

No, there’s only one solution:

1. Wide lander with low center of gravity

2. Slow careful vertical descent

3. Flat landing site

[MartGonzo]

I like infernal robotics for this problem, structural parts work just as well but pistons are fun

[*Aqua*]

It can help to rotate your lander on decent so one landing strut faces downhill. Right click on it and lock its suspension. If you have time you can also raise the uphill landing struts (and pray to god the lander won't fall over).

That simple technique helped me countless time.

[KerikBalm]

I've also experimented with designs that can tip over... by putting landing legs to stick out sideways. With better than a 2:1 TWR, it should be able to lift off.

But I found the landing pads to be kind of sticky, and it would kind of nose over. If only there was a way to put castoring wheels on the ends of the landing legs.

Othertimes, I've tried to asymetrically retract landing legs according to the slope below.

I like to use LV-N landers, with only a single LV-N (for surface gravities below that of Tylo), so my designs end up quite tall.

RCS ports at the top for maximum leverage... cubic struts/girders to widen the base.

Slow touchdown speed

Quickloads

[11of10]

I'll just drop this here... my first successfull munar lander

[VincentS]

you can actually land a tall (or upside-down cone-shaped) lander on a slope, provided that it has a narrow contact zone (e.g. single LVN without feet), reaction wheel (with SAS on) and a slow enough landing. But don't forget to bring enough battery !

If SAS is powerful enough with respect to gravity, your lander will look like a top and stay upright for as long as you have electricity. This obviously is easiest to achieve in low-gravity world.

[Brainlord Mesomorph]

I like to use LV-N landers, with only a single LV-N (for surface gravities below that of Tylo)...

would there be a problem w/ a 1 LV-N lander on Tylo?

(I'm working on one right now.)

Edited October 31, 2014 by Brainlord Mesomorph

[Zylark]

I must admit that for most my landings, I just use MechJeb that do a pretty good job of precise and safe landings. There are however a few cases where this is not the case. Early in career means doing it manually, and from time to time if landing on some new location, the spot you've picked is on a heavy slope and one need to take over the controls to do a manual landing somewhere nearby that is a bit more level.

So my landers are usually pretty squat designs, where a low CoM and wide separation of landing legs negates any minor horizontal drift as I touch down. I also tend to go with triangular design for landers, so that no matter how uneven the ground is, all landing legs will touch the ground. Quad designs may leave one leg hanging.

The short of it being - make stable designs and don't land on a heavy slope

[Laie]

would there be a problem w/ a 1 LV-N lander on Tylo?

(I'm working on one right now.)

No fundamental problem, it can be made to work; you may find it impractical, however.

Tylo has a relatively high gravity (0.8 Kerbin-Gs); an LV-N can lift only about 7.5 tons, and 2.25 tons of that is the LV-N itself. Add in the fuel for landing and takeoff, and it may well be that you can't even add a MK1 Lander Can. Certainly not much more than that: no legs, no science.

[Starman4308]

would there be a problem w/ a 1 LV-N lander on Tylo?

(I'm working on one right now.)

The lower the TWR on your lander, the more you lose to gravity drag. In the case of Tylo's beastly gravity well, this might become such a concern that one of the conventional 370-390 Isp rocket engines might be more efficient overall, due to the LV-N's poor TWR.

[bdito]

Wide base and MechJeb landings at 0.5 m/s.

[RexKramer]

would there be a problem w/ a 1 LV-N lander on Tylo?

(I'm working on one right now.)

I've used a two LV-N lander on Tylo with success. It was a fairly heavy lander. I'd imagine a single LV-N lander is possible if you keep it light. The problem with a single LV-N is keeping the center of mass low, because those engines are long.