What am I doing wrong when landing on the Mun?

[photogineer]

Hey there all, me again, I don't post much!

Last time I posted I asked for help with getting in to orbit, and I got a lot of great answers, so I'm doing it again. I took a break from KSP shortly after that and resumed when 0.22 came out on my new Mac, and I'm loving the game again.

So here's the dilemma, I'm trying to land on the Mun, I've watched every video there is, I've tried over and over using quicksave and I just can't quite get it. I'm close, but not quite there. One of my landings ended with a crashed pod full of Kerbals, so now I want to rescue them. Before I get to that too, I'm trying to master landing and making it back from the Mun without worrying about where on the Mun I land to pick them up. To that end, I've built and can fly a rocket (the Avenger II) in to Kerbin orbit, I can transfer to a nice 90km circular Mun orbit, but whenever I attempt to land it goes the same way. So, here is what I am doing:

1. From a 90km parking orbit I burn retrograde to drop my periapsis down to about 1km.

2. I ride out the descent, near the end I burn hard retrograde to null my horizontal velocity out, and this seems to use up a LOT of fuel!

3. I drop vertically, periodically firing the engine to control my descent rate.

4. Just before touchdown I fire the engine at a high throttle settle to slow my descent to a crawl.

5. I touch down, almost completely out of fuel, and then fall over and break. Every. Time.

So my big problem here is the tumbling over, but I'm landing on a flat surface, I've killed my horizontal velocity, my lander is completely symmetrical, I have no idea what I am doing wrong. Also, this method seems to use up too much fuel for me to return to Kerbin easily/at all. So I have to think I am doing something or things wrong here! Any advice?

Matt

Edited November 16, 2013 by photogineer
Changing to answered

[Scarecrow88]

Sounds like you are making a reasonable attempt at the landing, aside from running out of fuel and falling over. I usually start my descents from a much lower orbit, 30 - 40k but only because it makes it easier for my returning craft to rendezvous and dock with the bit that stays in orbit.

Do you have a screenshot of your lander, as it sounds like there might be a design flaw if you are landing with negligible vertical or horizontal velocity and it is still falling over. May also help to see what your fuel issues are as well. Alternatively, if you can host a link to your craft file somewhere, I'd be happy to download it and fly it to see if I can identify any problems with the design.

[Vanamonde]

Yes, that sounds more like a design problem than a piloting issue. Post pics of your ship? Low and wide is more stable, and don't use the medium-sized legs because they LOVE to flip your ship on its head.

[photogineer]

Design flaw eh? I'll buy that, though my designing skills I hoped were better than my piloting skills! (It's not as if they could be worse!)

Craft file is here: https://www.dropbox.com/s/ahqroa2g4h7zieh/Avenger%20II.craft

Here is a screenshot in Kerbin orbit of just the lander stage: https://www.dropbox.com/s/769ydrqj3s4bwig/screenshot171.png

Here is another perspect right after cutting the lander loose for reentry: https://www.dropbox.com/s/hba2ivmz1cv4bgp/screenshot173.png

Those are all the screenshots I have for the moment, I was too busy frantically trying to control it on the way down to take any more!

Matt

[Andersenman]

Just to make sure: When you land and the speed display on the navball shows comfortable numbers, it is set to "Surface", not "Orbit", right? Also, your landers might be too fat and too top-heavy, the legs too narrowly attached and your control authority (gimbal, reaction wheels) insufficient.

[...] and this seems to use up a LOT of fuel! [...]

One way to estimate burn duration and time-to-node for a sort-of efficient suicide burn is to place a manoeuvre node where the orbit goes into the map mode's Mun's surface and keep pulling the retrograde handle for long enough to cancel all orbital speed. It'll be a bit too much since the surface is travelling at some speed still, but it brings you close enough. Anyhow, this manoeuvre should save you some unnecessarily early burning, therefore reduce wasting fuel. If that's still not enough ... bring more fuel, and a stronger engine to reduce linger time - remember: You don't just have to kill your speed, you have to do it while gravity is pulling on your ship, trying to accelerate it. The faster you kill speed, ie. the shorter your burns can be, the more efficient your landings are by reducing the time you give gravity to work against your burns.

Hope that helps.

A.

Edited November 11, 2013 by Andersenman

[Vanamonde]

Yep, that is TALL lander. High center of gravity. Consider these designs as suggestions/examples:

[photogineer]

Just to make sure: When you land and the speed display on the navball shows comfortable numbers, it is set to "Surface", not "Orbit", right?

It is set for Surface yes, also I have Kerbal Engineer running showing my horizontal velocity as some blithering low mm/sec value.

Also, your landers might be too fat and too top-heavy, the legs too narrowly attaced and your control authority (gimbal, reaction wheels) insufficient.

I'd believe that, perhaps you can look at the screenshots and .craft posted above and let me know if you think that's too tall? I was trying not to need a separate lander/return stage for simplicity but maybe that's killing me.

One way to estimate burn duration and time-to-node for a sort-of efficient suicide burn is to place a manoeuvre node where the orbit goes into the map mode's Mun's surface and keep pulling the retrograde handle for long enough to cancel all orbital speed. It'll be a bit too much since the surface is travelling at some speed still, but it brings you close enough. Anyhow, this manoeuvre should save you some unnecessarily early burning, therefore reduce wasting fuel. If that's still not enough ... bring more fuel, and a stronger engine to reduce linger time - remember: You don't just have to kill your speed, you have to do it while gravity is pulling on your ship, trying to accelerate it. The faster you kill speed, ie. the shorter your burns can be, the more efficient your landings are by reducing the time you give gravity to work against your burns.

Hope that helps.

A.

I'll give that a shot next go around too.

Matt

[Andersenman]

I've just seen your craft's image. Clearly, you are too fat. No, seriously, you bring too much fuel, fuel mass you have to decelerate. And once you're down, your craft is very top-heavy (full RCS, I figure), amplifying even slight imbalances. You should easily be able to land and return with what a Kerbal X provides on Mun - which is still 2.5 km/s dv and a TWR of 8-point-something with a third of the fuel and tank weight you bring. Also, you should strongly review how much RCS and reaction wheel power you really need and ditch the rest. So, against my original advice: Bring LESS fuel.

I'll give that a shot next go around too.

Don't forget that the usual half-duration-before-half-duration-after-node thing doesn't apply here because, well, the second half of the burn would be spent below ground.

Happy landings!

Edited November 10, 2013 by Andersenman

[codepants]

Well, you can land anything with legs, even if it's tall, if you land on a flat surface and come down slow enough. Even if I don't have to, I always try and touch the ground at < 1 m/s. It's hard sometimes, but it will greatly reduce the chances of your craft tipping over.

Also, once you've gotten rid of all your horizontal velocity, you can set SAS on when pointed straight up. Then if you start to tip, it will (try and) keep you upright. If the ground isn't flat, once it's eliminated your roll velocity, tap F to adjust to the ground a bit more. On smaller moons you can just have it keep your craft pointed up while you go take surface samples and such though, it's a lot less risky.

Good luck!

[capi3101]

Yeah - what you've got there is a variant on a "Phallus 7" (if you'll forgive me for being crude) - a long and narrow rocket with a very high center of gravity. Most folks design a lander like that (Chute, Mk-1 Command Pod, FL-T400 tank and LV-909 with 3-4 lander legs) for their first trek out; this one's larger of course, but essentially the same idea. The problem is in the design; it's easy to tip this kind of rocket over.

Here's how I'd go about fixing it:

1) Dump the big RCS tank - I'd wager you're not using that much RCS anyway. If you're not comfortable with doing away with RCS completely, I'd humbly submit you use 4-6 of the small Roundified tanks (read: I recommend zero, two at the most if you must have some). You'll save mass (and therefore improve your delta-V).

2) Consider swapping out the Mk1-2 Command Pod for three Mk1 Lander Cans. Again, there's a mass savings there.

3) Here's my key suggestion - swap out the X200-32 tank for four X200-8 tanks, and set each one radially outboard of the X200-16 as low as you can attach them. Place your lander legs on those tanks, and run fuel lines from those outboard tanks to the central tank. That widens and shortens your stack while keeping the same amount of fuel.

I'd also humbly suggest getting closer to the Mün's surface before you burn to deorbit. You can safely orbit as low as 14 km above the surface, and you won't pick up as much velocity in the process of your deorbit. You'll still have to burn a good deal to bring your horizontal velocity to zero (you've got a heavy ship there, that's unavoidable), but it should save you some gas.

You only need about 850-1000 m/s of delta-V to make it back to Kerbin from the Münar surface (assuming you acheive Münar orbit first). Assuming you dump the RCS tanks, your "bingo fuel" level is therefore just over 200 liquid fuel units; if you go below 210, hit space or plan to not return from the Münar surface with that flight. With the RCS tank, "bingo fuel" is 300 fuel units. If I'm doing the math right, you start with 2430 - so it won't look like you've got that much left either way. Good news is that you don't need all that much.

Edited November 10, 2013 by capi3101

[Vanamonde]

Well, you can land anything with legs, even if it's tall, if you land on a flat surface and come down slow enough.

Yeah, but some configurations are easier than others, and flat spots are hard to come by on Mun's new terrain. And that's assuming that what looks like level terrain when you're coming down from above actually turns out to be flat when you get there. You're better off building a lander that can handle the terrain you find than trying to find terrain that can handle the lander you've built.

[Andersenman]

You're better off building a lander that can handle the terrain you find than trying to find terrain that can handle the lander you've built.

Out of interest, apart from the obvious things like low CoM and wide base, how DO I build a lander in KSP that can handle anything but reasonably level ground? It's not like KSP's legs are coded to equalise slopes and keep the craft vertical (which would be really awesome by the way).

[Spaceweezle]

@Vanamonde - Those landers of yours are too cute! Nice designs man.

[Kasuha]

Make it short and wide with enough legs and you can land almost anywhere.

Having the ship vertical and SAS and RCS on just before touchdown helps a lot to let the ship turn to the slope without falling over. Also be gentle with throttle, turning it off abruptly is not always the best idea, but you need to bring it down before you start sliding down the slope.

Edited November 10, 2013 by Kasuha

[photogineer]

Alright everyone, thanks for the tips, I'm designing the Avenger III now with a better lander configuration, I'll post here with my progress!

Matt

[Spaceweezle]

Alright everyone, thanks for the tips, I'm designing the Avenger III now with a better lander configuration, I'll post here with my progress!

Matt

Please do because I like pictures!

[KSK]

One other thing that might help with the fuel efficiency is to tweak that landing profile a little. I tend to set a periapsis of around 7-8km for my de-orbit burn and then start my landing burn at periapsis, going for a single continuous burn if possible and ideally. zeroing out my horizontal velocity about 1km from the surface, so I'm not burning too much fuel for that last descent. It rarely works out so smoothly but that's what I aim for.

I find I don't have to do a lot of maneuvering until I'm down to the last 150m/s or so. At that point its time to throttle back and follow the retrograde marker.

[Vanamonde]

apart from the obvious things like low CoM and wide base

Apart from those? Not much, though RCS can sometimes prevent tipping.

Those landers of yours are too cute! Nice designs man.

"Cute"? CUTE?! There is no cute when lives are at stake! (And thanks. )

[Taki117]

I did a Kerpollo style mun landing if you want ideas on how much fuel I brought with me. It was enough to get there and just enough to get back. I can provide a craft file if you need. Here's the mission report:

http://forum.kerbalspaceprogram.com/threads/57652-Breaking-news-Kerpollo-to-the-mun-and-back%21

[Specialist290]

This tip applies more for the design as a whole rather than the lander specifically, but one thing I try to do is break the mission down into phases and design separate components for each phase. Specifically for Mun missions, I'll often break it down like this:

1. Launch to LKO, achieved by a dedicated launch vehicle that is discarded once I'm in orbit.

2. Trans-Mun Injection and circularization in Mun orbit, done by an insertion stage separate from the lander itself. (If I have fuel left, I'll sometimes start the landing phase with whatever's left in this stage, to give my lander a little more "elbow room.")

3. Landing and return. Depending on what I feel like doing on that particular day, I'll either have an "all-in-one" direct ascent lander, or I'll make the lander and crew return vehicles separate, Apollo-style. Either way works well in KSP, although the latter is a slightly more complex design challenge by the numbers.

Designing each component to perform a separate task also gives me a good measure of whether or not I can accomplish my mission based on how the vehicle performs. If, for instance, I'm desperately burning my insertion stage to get a stable orbit around Kerbin, or if I have to use my lander engines to circularize at the Mun, then I know that I've messed up somewhere and will likely need to start over again.

[700NitroXpress]

If you have a really wide landing base, then it will be almost impossible to flip a lander over. You also want your lander to be bottom heavy and not top heavy, as top heavy craft will flip over and a bottom heavy craft will be really difficult to flip over. As far as fuel consumption goes, use a lander that's lightweight with small engines that give it a huge TWR. You can make a lightweight lander capable of getting into Kerban orbit all by it's self with very small engines and a compact design. Do that and Mun landings are a piece of cake.

Here a Mun landing and return video I did a while back, see if it helps with your attempts, this is a heavier lander, but it works well for landing anywhere.

[photogineer]

Please do because I like pictures!

Here are some! The new lander is much more compact, but the launch vehicle needs more work. The first prototype I launched also lacked parachutes so after a successful Mun flyby I was left in a wild elliptical Kerbin orbit with no chutes and low fuel! I tweaked the lander stage to have chutes, played with the staging a little and sent up a rescue ship with the next revision. I was able to rendezvous with the first Avenger III capsule, only to realize I forgot to de-crew the ship and only had one open seat. So I brought Jeb home since he seems like the kind of guy who will be able to help rescue the rest of them! He booted Milny out of the capsule in to the lawn chair strapped on top, and Milny rode through reentry in that!

Here's the current launch vehicle on take off: https://www.dropbox.com/s/qrbvok8k0gcb14z/screenshot225.png

Here are some photos of the lander capsule (w/o parachutes):

https://www.dropbox.com/s/r7d1jwfenp72pbt/screenshot211.png

https://www.dropbox.com/s/w6p8q9jt5zf22yl/screenshot212.png

https://www.dropbox.com/s/fkw3794za640v43/screenshot213.png

Here it is with them: https://www.dropbox.com/s/ex2mfzmfbg6e08x/screenshot246.png

The rescue ship approaching the first prototype: https://www.dropbox.com/s/7ylgxynsowhmnoc/screenshot251.png

Parked next to them (this is when I realized it was fully crewed and started cursing!): https://www.dropbox.com/s/wk7niuzxgyp43j1/screenshot254.png

Jeb getting out after ordering Milny in to the chair: https://www.dropbox.com/s/flhupu2jhfkajou/screenshot260.png

Rescue ship rocketing off to land leaving some men behind: https://www.dropbox.com/s/vcmqj0dljyjrt2p/screenshot266.png

Aiming to land in Booster Bay: https://www.dropbox.com/s/2lof872grtgwjcg/screenshot272.png

Missed it! Oh well: https://www.dropbox.com/s/1dbebl0vb186tul/screenshot279.png

Milny getting really warm and toasty (Jeb thinks it is hilarious): https://www.dropbox.com/s/ucbso4p9fqbdv2d/screenshot285.png

Chutes out, gear down, we're coming in over land! (Jeb really can't aim can he? Should have picked up Bill instead...): https://www.dropbox.com/s/tax6qy965rq4ahu/screenshot290.png

Milny just wants to be home, with snacks: https://www.dropbox.com/s/ofabyq7hnblh3ia/screenshot295.png

Jebs firing RCS as makeshift soft landing jets, they don't help at all: https://www.dropbox.com/s/baqam1n0ac9aqnj/screenshot302.png

Turns out the landers engine extends below the lands when they compress, this was terrifying: https://www.dropbox.com/s/wpx5smsea9w1u1g/screenshot304.png

But they're all OK!: https://www.dropbox.com/s/gicdch28n5xn7jx/screenshot305.png

Here are the boys awaiting pickup, at least the lander landed on something right: https://www.dropbox.com/s/4tlmm8rxghcpnn0/screenshot308.png

Their final landing position:

https://www.dropbox.com/s/1t76qx6t9fml3aj/screenshot309.png

https://www.dropbox.com/s/qcdgw01mzr1ikke/screenshot310.png

It's progress! Even though I somehow ended with 2 extra Kerbans to rescue...

One other thing that might help with the fuel efficiency is to tweak that landing profile a little. I tend to set a periapsis of around 7-8km for my de-orbit burn and then start my landing burn at periapsis, going for a single continuous burn if possible and ideally. zeroing out my horizontal velocity about 1km from the surface, so I'm not burning too much fuel for that last descent. It rarely works out so smoothly but that's what I aim for.

I find I don't have to do a lot of maneuvering until I'm down to the last 150m/s or so. At that point its time to throttle back and follow the retrograde marker.

Are you saying you don't kill your velocity separately but do one long velocity kill and descent burn? Is this any harder to balance?

This tip applies more for the design as a whole rather than the lander specifically, but one thing I try to do is break the mission down into phases and design separate components for each phase. Specifically for Mun missions, I'll often break it down like this:

1. Launch to LKO, achieved by a dedicated launch vehicle that is discarded once I'm in orbit.

2. Trans-Mun Injection and circularization in Mun orbit, done by an insertion stage separate from the lander itself. (If I have fuel left, I'll sometimes start the landing phase with whatever's left in this stage, to give my lander a little more "elbow room.")

3. Landing and return. Depending on what I feel like doing on that particular day, I'll either have an "all-in-one" direct ascent lander, or I'll make the lander and crew return vehicles separate, Apollo-style. Either way works well in KSP, although the latter is a slightly more complex design challenge by the numbers.

Designing each component to perform a separate task also gives me a good measure of whether or not I can accomplish my mission based on how the vehicle performs. If, for instance, I'm desperately burning my insertion stage to get a stable orbit around Kerbin, or if I have to use my lander engines to circularize at the Mun, then I know that I've messed up somewhere and will likely need to start over again.

I am trying to do this, and not doing so great at it, so I am going to focus on componentizing my design more. Currently I have to use the insertion stage to finish the entry to LKO, so I know that's not working out in my favor. I'm getting closer though, I'll keep everyone posted!

Matt

[Andersenman]

Not to mention that it costs far less dv to bag a low arrival periapsis during your transmunar burn than lowering it after insertion into the first best high munar orbit you come across. And as usual, manoeuvre nodes are your friends, together with the different conics modes to visualise trajectories, so switch between them to get a feeling for them using any of the node-related add-ons. It's really not that complicated, don't overthink it.

Edited November 10, 2013 by Andersenman

[Scarecrow88]

This thread has grown since I first replied, and I see you have already made changes to your initial design, so my observations may be a little redundant or have already been made by others.

Your lander is way too tall to be stable and would struggle to stay upright on all but the flattest of landing sites. As you have no docking ports on the vehicle, I assume you don't intend to use this craft to rendezvous with others, which makes the RCS an unnecessary addition. There are no ladders on your lander. This isn't too much of a problem on Mun or Minmus, where the gravity is low enough for any Kerbals on EVA to use their jetpacks to get back in the capsule, but it could be a problem on some other bodies. The chute you have attached is a drogue chute, designed for atmospheric braking. I've never used one to land a returning capsule back on Kerbin, so am not sure if it has enough drag to do so successfully.

Having removed both the RCS and the larger fuel tank from your lander, it is capable of landing on the Mun but not with enough fuel to get back to Kerbin. This is principally because I had to use the lander to circularise the orbit at Kerbin and fly the Munar injection phase. If you can incorporate a more fuel efficient middle stage to perform these tasks, I am sure your lander will have more than enough fuel to perform a Mun landing and return to Kerbin.

Hope that helps.

[SpacedCowboy]

I would say your biggest problem is ( step 1. ) As others have suggested, try your retrograde burn from a lower (parking orbit) like 35 or 40