Landing on the Mun

[Vasudan]

So, after a few days of learning to get into space, obtain an orbit and slingshot to the mun, I\'ve been beaten by the actual landing everytime.

The problem seems to be the small amount of sideways movement coming down, I simply can\'t maintain a perfect vertical decent. Although I can be almost 100% vertical most of the way down, I always get some sideways movement in the last few hundred meters.

The result is the command module ever so slightly detaching from whatever happens to be left. It doesn\'t actually fall off, just breaks a tiny bit and sits slightly to one side, although technically a succesful Mun landing, kerbals alive and well, it\'s no longer connected to the boosters and the\'re never coming back. This is assuming the the RCS tanks don\'t explode leaving the command module all alone, stranded.

I\'ve tried a couple of lunar module mods with the same resut.

Am I doing something wrong or is it simply that fragile? My attempts at last second corrections to reverse the sideways movement usually result in firery explosions.

Would appreciate some tips.

Vas.

[BillWiskins]

As someone who has recently caused Mun landings to become fairly routine (endless, painful repetition and having a good solid pole nearby to smash one\'s head upon both seem to help), I\'d say these things:

How are you landing (RCS, liquid engine)?

Where are you landing? (Those lovely craters are a lot smoother and more level).

How are you correcting the lateral movement? It\'s a lot easier with evenly-spaced RCS than with one gimballing engine.

I\'m hardly the expert you want (and there are very many threads here which will help you more) but I\'ve managed it, with stock parts, and compared with many on here I suspect I have the piloting and construction skills of a cheeseburger.

Good luck!

[Vasudan]

Unfortunately my skills don\'t go so far as to pick my landing site, I\'m coming down where ever... but it\'s not the terrain that seems to be a problem when your drifting sideways, I\'ve landed smooth as, slid a bit and \'clunk\' off comes the command module. I have some screenshots of the lander pefectly touched down and the command module lying in the dust beside it.

I\'m using what standard liquid rocket/fuel I have left on most of the descent and after dropping that, I\'m sually 99-100% vertical. I had about 10 odd RCS spaced evenly around 2 RCS fuel tanks, but having found that way too fragile for landing I switched to a lander mod that has it\'s own fuel and is slightly easier to move around. It\'s that last couple of hundred meters when I\'m trying to wipe off the speed that I run into problems.

I think I\'ve just got to get better at the finer last minute adjustments and avoid the last second panic fireballs, when down to the command module those small thrusters are just sooooo sensitive.

Vas.

[BillWiskins]

I\'d say practise is a large percentage of it...I\'ve always used just the standard parts so I can\'t comment on flying a pre-built lander, but I can\'t think of anything outrageously, glaringly wrong with what you say you\'re doing, so I reckon just keep at it.

I guess it\'s not exactly uncommon or a secret technique, but when I\'m getting near to the ground I find it helps to pan the camera all the way over the top of the lander, so you are looking vertically down. This way it\'s very easy to see which way you need to thrust to kill any sideways movement nonsense that may be occurring.

[Kosmo-not]

I look at my navball to see if i need to correct any lateral movement. It helps to move the camera around to help you judge distance to the surface. I reckon I should make my own Mun landing video sometime.

Once you know how to do it, and do it a few times, Mun landing becomes a piece of cake.

[Hikaru]

There are a lot of little things I do that make mun landings easier. For one, while in HIGH orbit, I deliberately slow my movement to as close to zero as possible - this reduces (almost) any sideways movement to manageable amounts. During this phase I am (ab)using the map to determine where I\'m going to land - if I stack the center of the mun on top of the center of where the icon for my craft is on the map, I get a reasonably good idea of where my ship\'s going to land. I want to land on the lit side of the mun and somewhere in a crater, preferably. If I\'ve truly screwed this up, corrections can be made on the way down - but I\'ll need to go through a zeroing of the movement again, which will waste fuel - so it\'s a good idea to get it right the first time. When I\'m sure I\'m going to land, I point my ship vertically, turn on the rcs thrusters with an asas and try to drop down vertically the rest of the way without any further input. Sometimes this doesn\'t work and I have to make adjustments, so I again zero all my movement. When I get close enough to the surface to see the ship\'s shadow I should be moving at less than ten meters per second, better to have it even slower than that if possible. When about to touch the surface I want to have the ship\'s throttle set that the ship\'s thrust is being very slightly overcome by the gravity, so when I hit the ship won\'t suddenly bounce off the surface. Speaking of which, remember how I said I have an asas with rcs thrusters enabled all the way down? Three reasons for this: Any stability problems on the way down will be corrected automatically, when you hit the surface if you *do* bounce off, your ship will automatically recover if it decides to flip over, and if you *don\'t* bounce off the asas will assist to slow your ship\'s horizontal movement down since you will tend to tip towards the direction you\'re moving when you hit the ground.

And of course, the best way to learn how to land a ship is to do it. Make a nice sturdy ship with plenty of fuel and give it a shot.

[Kosmo-not]

Here. I said I would.

KSP: Mun landing tutorial

[E. F. Kranz]

Just keep the craft aligned with the ? during your descent. Don\'t even look at the ship, just fly by instruments. Watch your descent rate and altitude, but other than that, fly on the yellow dot. You will find as you do that it moves closer to straight up. Follow it. Don\'t be distracted by the ship.

[Red Iron Crown]

I zero out as much horizontal velocity as I can while still 5km or more above the Mun, using the map view to get the orbit to look like a vertical line. Then I set the lander\'s nose straight upward, turn on RCS and SAS and adjust the main engine to give an acceptable sink rate.

As I get closer to the surface and can more accurately see my lateral speed, I use the translate keys (I,J,K,L) to use RCS thrust to get me dropping straight down, then a couple of quick hits of H just before touchdown to almost zero the vertical speed. Then hit X as you touchdown to kill the main engine and settle the lander.

It can be stressful, as you juggle throttle position, attitude and translation with the surface coming up at you. Definitely gets the heart pumping!

[Xivios]

After some trial and error (mostly error) I\'ve finally constructed a moonshot rocket, entirely from vanilla parts, than can deliver, land, and return a crew from the mun with a relatively high success rate - indeed, it has plenty more fuel than it really needs.

Here is what I\'ve found.

You\'ll need a sturdy vessel. As far as I can tell, the best 'lander legs' in the vanilla game are the movable fins - they dip far down at their trailing edge wingtip , well below where they attach. The fixed fins don\'t, being a trapizodal shape, so though you might not want your legs to wiggle about, you may have to live with it unless you mod or Harv and the boys add some standard legs. Don\'t worry, their collision mesh does not move, so the movement does not affect their use as legs.

They are very low mass, and also have the highest collision rating of any part, so do not seriously affect the rocket. They do make it much more difficult to get out of the atmosphere, due to their drag, and their location on top of the rocket, when used in this role, is not conducive to stability. Combine them with an Advanced SAS though, and they manage to pull double-duty though on the stability end of things though. I don\'t touch anything but T, Shift and Space until 30,000 meters on me rocket.

You can struss them up with struts too, they must always be the second item attached though. I like 3 struts each, one left, one right, one above. I use 6 fins total, 6x symmetry. Plenty of strength results, the thing can slide (and has) down hills on Kerbal and not loose a fin (well, it lost one). It\'s also fairly big - my lander is comprised of: 1 gimballing LFE engine, 7 fuel tanks, 2 RCS fuel tanks, 12 RCS motors, 1 Advanced SAS, 6 fuel transfer lines, 6 moving fins, 24 struts, 1 stack decoupler, the command module and parachute. I think this size helps it a lot - the high inertia acts to damp the control inputs and makes it all smoother. This is actually too heavy to lift off the ground with full tanks on Kerbal, but it plenty potent on the Mun.

So, usual approach then is pretty close to E. F. Kranz\'s method. Fly by instruments, beyone checking that the ground is reasonably level and using shadows (if not eclipsed by Kerbal, it\'s happened to me) to gauge final descent distance.

Actual approuch and landing is as follows.

After entering moon SoI enter moon orbit between 10,000 and 5,000 meters.

On approach to desire landing site convert velocity readout to Surface relative (not orbit), point ship retrograde and fire main engines. Holding ship with the retrograde marker (?) centered in the navball, much as E. F. Kranz\'s method suggests, and keep it there until the ship approaches 0m/s. The retrograde marker will rapidly move just before this occurs. When it rapidly moves out of the way, press X to kill the engines. You have probably now managed to reduce your forward and lateral motion to around 3m/s or less - indeed, with some finess you could probably attain a velocity of about 0.3m/s before killing the engines - this is far more accurate than Red Irons map method, and will place you into a nearly-perfect vertical descent.

Once the ship begins to freefall vertically, re-orient the ship vertically and place the retrograde market back into the center of the navball.

Try not to burn too much fuel doing so but manage your descent speed to what you feel is safe.

On approach below 500m indicated altitude, reduce speed to about 5m/s and hold it around there with careful throttle control. Try not to go too slow - doing so will exagaggerate the lateral retrograde markers movement - remember, at 5m/s, a lateral velocity of 0.1m/s won\'t perturb the marker too far from center. At 0.1m/s however, with the same lateral motion (ie, all of it now) the retrograde marker will be completely on the side, and you will probably overcorrect for it - and that\'s where the shit hits the fan - but continue to follow the marker gently, keeping speeds lower and lower until touchdown, figure around 3m/s at touchdown. Then x to kill it. The total velocity at touchdown is more important than your lateral velocity, and if you follow the retrograde marker smoothly and accurately, it\'s entirely possible - and sometimes easier (because it\'s more conventional controls) to do so without RCS - but RCS can help so don\'t shy from it if you want to use it.

Edit: Oh, yeah, if things start to go sideways, figurately or literally, always remember: Burning directly retrograde, with the marker center to the navball, and doing so as perfectly as you can until speed is as close to 0m/s as you can manage and then killing engines will always put you back into a perfect vertical freefall. Don\'t be afraid to do so if you need to straighten things out. Just remember that the marker will go wonky-like when you approach 0.

[Newman]

In short, these guidelines should be of help:

- make your lander as light as possible. The larger/heavier the ship, the harder it will be to land. The trick is to balance the lander so you have enough fuel to land and return home again without bringing too much mass to the moon. But you\'ll find it easier to land with a lighter lander, when you get the hang of it, you can try the more convoluted designs.

- if you\'re not a 100% devoted purist, use lander legs. They make it much easier.

- you pick your landing site from Mun orbit. If you don\'t establish a proper orbit prior to landing, you\'re doing it wrong. Look up Ivan Ivanovich\'s excellent guides on this to gain an understanding of basic orbital mechanics, and how to modify your periapsis, apoapsis, and inclination. Setting up your orbit so it goes over a crater is easy. After that you just need to break at the right time. Use the map, and you should be fine

- when you\'re ready to begin your deorbit burn, try and get as close to zero speed as you can by following the reverse vector icon on the navball. Have plenty of altitude when doing this - around 60, 70km altitude works great. This should more or less result in an almost straight line down to the surface. It\'s easier in a crater, but I\'ve managed landings pretty much on every type of terrain.

- control your descent so you don\'t get so much speed so you can\'t break in time. Look at the vector icon on the navball, and try and keep it dead center.

- when at a lower altitude, you should have cut your velocity to a manageable one. Use SAS to lock your orientation dead center, use your main engine to control vertical velocity, and use the RCS translation controls to control horizontal velocity. Though I\'ve done landings without RCS at all, it\'s easier like this, especially on a smaller lighter lander the RCS will have more of an effect on.

- keep your vertical velocity as low as possible and prepare to kill engines completely the moment you touch the ground. Anything below 5 m/s should work. My best landings managed to get speed almost killed completely before landing, without any horizontal velocity - you just gently touch the surface.

[4 IN 1]

Well you will have to learn to land with heavy payload on the moon if you were to build a moon base, so better early then late =P

[Awaras]

Coming to a velocity of zero in orbit does not seem very fuel efficient to me, correct me if I\'m wrong...

I found that this trick helps me a lot to land on the mun with zero lateral velocity, and I almost never need to use RCS:

When you turn your ship retrograde but move the ship a bit to the left or right so that the retrograde icon is not precisely centered on the navball, as you apply thrust the retrograde icon will start moving directly away from the center of the navball.

As you descend, you position your ship so the retrograde icon is located directly between the center of the navball (your 'heading') and the little dot on the 'north pole' of the navball (in the center of the blue navball hemisphere). As you thrust, the retrograde icon will move towards the center dot. When the retrograde icon aligns with the center dot, you have zero (+/- 1 or 2 m/s depending on your precision) lateral velocity. Then you align your heading with the same dot and keep thrusting down to land.

Of course, you do all this while you are descending, slowly decelerating all the way, slowly moving your ship towards the vertical. I usually put myself in a stable orbit around the mun, wait until I am in a position to land where I want and then decelerate just enough so that my orbit intersects the desired landing spot. I start to burn off lateral velocity as described when i reach an altitude of 50-100 km, depending on the acceleration of my ship.

[Elukka]

This type of lander is the easiest to land:

It has comparable thrust to a full liquid engine and a bunch of fuel tanks due to it being much lighter. You can also stay upright for the final descent and use RCS to zero out your lateral velocity. It can also return, though you\'ll need a burst from the engines before Kerbin landing to avoid exploding.

[Rickenbacker]

Coming to a velocity of zero in orbit does not seem very fuel efficient to me, correct me if I\'m wrong...

It can be, if you then fall almost all the way and slow down at the last moment, or close to it.

I find that using ASAS helps. I use it to keep my lander pointed straight up, then I use the yaw and pitch keys to point the exhaust the way I\'m moving, to slow down and eventually come straight down. That\'s the method that works for me, but it makes for a heavier lander, so it might not be ideal.

[ash73]

I started out landing by eyeballing the terrain from above, but I switched to flying off the instruments and it\'s much easier. I just keep the little green circle centered and follow it all the way down, with a quick look out the window to check for dodgy terrain. Above all else, don\'t panic! If it\'s not looking good just bale out into a hover, settle down and try again.

[closette]

Some good approaches (pun intended) given in this thread. In terms of fuel efficiency for a Mun landing on a given spot from an arbitrary position and velocity for a given spacecraft, that\'s a complicated 'optimal control' problem.

Here are some relevant links:

http://en.wikipedia.org/wiki/Gravity_turn - scroll down to 'Descent and Landing Procedure',

and for some light reading: http://trs-new.jpl.nasa.gov/dspace/bitstream/2014/41890/1/11-0208.pdf - 'Optimal Terminal Descent Guidance Logic To Achieve a Soft Lunar Touchdown'.

I\'m usually just happy to land safely and not too fussy about picking a spot, although Kosmo-not\'s video was very instructive. It looks like a ~10 km orbit is the lowest you can safely go for a 'drop down from orbit' technique, but I am curious myself if it would be more, or less, fuel-efficient to drop down from a higher orbit? (You\'ll have less orbital velocity to cancel out, but farther to fall vertically afterwards).

[Zephram Kerman]

Wow! This is a very helpful thread. Thanks to everyone for contributing. Until just now, I did not know about the RCS translate keys I J K L. I\'m sure that will be handy from now on.

There\'s just one thing remaining on my mind. I don\'t see how anyone manages to pick a good spot. The big craters are said to be nice. But even in the tutorial video, as we get closer, steep slopes appear out of nowhere.

Have I missed a graphics setting, or something? Alternatively, I\'ll just hover and cruise around with RCS translate.

Why is the VCR flashing 1201?

[Xivios]

You\'ll never find a really flat spot on the Mün, just sort of wing-it and hope for the best. As long as it\'s not too steep, you\'ll be OK. I\'ve landed on some pretty nasty terrain - on Kerbin, no less! The lower gravity of the Mün makes it even less likely to break things apart, as you as you touch down gently.

If things still look a bit nasty to you, well, there\'s a reason I like bigger landers. More fuel, gain some altitude, boot it over a few hundred meters, perhaps into the valley of hill you don\'t want to land on, and try again.

[Zephram Kerman]

Sooooo...

Roomfuls of the best and the brightest poring over hi-res stereoscopic photos made by earlier scouting missions at great expense? That\'s not Kerbal.

'If you see slope, punch it.' That\'s Kerbal!

[semininja]

I\'ve found that with a deorbit engine on the lander, I can get down to about 1km orbit before landing. Also, the mare (large craters, darker than surrounding surface) may be flatter than the rest of the surface, as on the real moon. I\'ve been using this to help pick landing sites, so you may benefit from it.

[Zephram Kerman]

Thanks Semininja, that helped a lot too. Also, it\'s kinda fun to soar over the munar terrain. I did some sightseeing before going in.

(Crashed, of course. But it was on level ground this time. Yay!)

[Drakomis]

I really love this thread. I read up on some of the tips and I have to say, it\'s made my trips and understanding on how to land a whole lot easier. Good job on everyone that contributed. Also, thank you very much.

Your friendly neighborhood bag of bones,

-Drakomis

[closette]

Well I am going to answer my own question above: for landing from circular orbit using the 'stop n\' drop' technique (slow down, then fall vertically), I found the impulse for each part assuming an idealized model, where the impulses are instantaneous (i.e. thrust can be infinite). I am also assuming that the change in spacecraft mass is negligible over the whole maneuver.

Starting at a distance r from the planet\'s center, the delta-v to stop the circular orbit is Sqrt (GM/r).

The delta-v to cancel out the velocity acquired after falling from that distance r to the Mun\'s surface at R=200 km is given by

Sqrt(2GM(1/R-1/r). This comes from energy conservation, if you\'re wondering.

Adding these, and noting from the Wiki that Sqrt(GM/R)=570 m/s for the Mun, the overall delta-v required to land is then:

570 m/s (1/Sqrt(x) + Sqrt(2-2/x)) where dimensionless x = r/R.

The smallest sensible value of this function is at x=1, i.e. r=R or zero altitude (!), and it has a maximum at x=3,

i.e. r=3R, or altitude r-R =2R=400km. Graph attached.

That surprised me at first, I thought there would be more of a trade-off between orbiting (slower with increasing altitude) and falling (faster with increasing altitude). But the short answer is, the lower your circular orbit, the less fuel you will need to land. Of course in practice you\'ll need to be at least a few thousand meters up because unlike my simplified model, it takes time to slow down. And if you\'re higher up, you can see landing sites ahead better. I think Kosmo-not\'s video above shows a good compromise, starting at 10 000 m circular orbit and dropping down from there.

[Iskierka]

That model also does not consider another factor: the delta-v required to get to that orbit. From a theoretical infinitely distant approach point, while the highest efficiency descent is from a low orbit, that orbit required more delta-v to enter in the first place. If you can enter a free-return trajectory of 10 km, circularise, and descend from there, it would certainly be more efficient than attempting to move up and descend, but if you entered at 100 km? Considering the reduced requirements for circularization, it\'s entirely possible it might not be any more efficient to descend to 10 km before making your final approach. Could be something to factor in there.