Mun Landing Precision

[The Space Dino]

Recently I landed a rover on the Mun to start building my Mun base, and of course most Mun bases have multiple modules. To minimize the time to drive the rovers to the modules in REAL TIME, I tried to land the parts as close to each other as possible. For me, I landed my second Mun base rover about 8 km from my first rover and totally screwed up the drive by flipping over and destroying my solar panels and lights.

What about the precision on your Mun landings?

*If you wonder how I got the second rover to the first rover, I did use Alt+F12 

[Streetwind]

Well, I don't build surface bases, but with a little practice it shouldn't be hard to consistently come down right on top of an intended target. You get used to it quickly when trying to visit anomalies

It's easiest if your base is right on the equator, or right on a pole. If it's halfway up a hemisphere, you have to contend with the ground slipping through underneath your trajectory as you attempt to land, and you'll have to learn to properly lead your target.

[Superfluous J]

Basically what Streetwind said. You need - more than anything else - practice. I used to be happy when things were a dozen km apart. Then I was mad if they didn't load into the physics window. Now I am more concerned with not hitting the base's solar panels than I am landing too far away.

Just keep it up. Try to figure out why you are missing and do your slowdown burn earlier or later. Or make sure you're heading toward the base exactly so you don't miss it to the side. Target the base, and make sure your prograde is heading toward (but over, as gravity will pull you down slowly) the base as close as you can make it, and correct. Get used to how the camera pans as you come down, tricking you into thinking you'll miss (or hit) your target, and learn to compensate.

[KerikBalm]

Set as target and using the nav ball to see where the surface velocity vector is pointing relative to the target helps a lot.

A high TWR also helps for when you find yourself overshooting and need to slow down fast.

 

This is the result of 3 "precision" landings. The first one was just meant to be close to the big ISRU in the background, and it deployed 1 module of the servive bay, then it went to orbit and took down a 2.5m module 2 more times so they could join together to form the surface base.

Those 2.5m modules have a very narrow track, and the wheel motors are weak so if most wheels aren't touching, it can't really climb a slope... they are not goo at all for roving long distances. I think I always put them down within 100 meters of the target.

Edited September 23, 2016 by KerikBalm

[marcello639]

What KerikBalm said. By using navbal and targeting I had situation where I had to change direction in very last second to avoid landing on top of the other lander.

[The Space Dino]

Hmm.... you guys are quite good at this (with practise of course) 

Any people who have epic fails to share?

[Guest]

I built one biggish Mun base in one career. Got to the point where I can get to within a few hundred metres no problem. Still a long way from being able to pick up the components with a skycrane (would need to align the docking ports on landing), but I've no doubt that's doable too, with sufficient practice.

[The Space Dino]

3 minutes ago, Brikoleur said:

I built one biggish Mun base in one career. Got to the point where I can get to within a few hundred metres no problem. Still a long way from being able to pick up the components with a skycrane (would need to align the docking ports on landing), but I've no doubt that's doable too, with sufficient practice.

I thought of using a sky crane too but I had no idea how to build or launch it or fly it.

[Guest]

5 minutes ago, The Space Dino said:

I thought of using a sky crane too but I had no idea how to build or launch it or fly it.

Skycrane deliveries are pretty easy: you just use a docking port or stack separator with decoupling force set to zero (or near as) between the surface module and the skycrane module, and put radial engines on the skycrane, if necessary angling them out a bit so the jets don't melt the payload (Twitches are good, and Spiders work for very low-grav bodies/very light payloads; Thuds work for very heavy payloads). You land them the usual way, decouple after landing, and fly off with the skycrane.

It's even possible to deliver a stack of modules this way – decouple the bottom-most one first, then take off with the second, deliver that, repeat until out of modules – although if the stack gets very tall things get hairy as it's likely to tip over.

In theory it ought to be possible to recover modules with a skycrane as well, but that would require precision piloting that's beyond my abilities at least, as you'd have to land on the target module's docking port to a precision of tens of centimeters.

[pandaman]

23 minutes ago, The Space Dino said:

Hmm.... you guys are quite good at this (with practise of course) 

Any people who have epic fails to share?

I think anyone who's ever tried it will have an epic fail or two.  I've had so many I can't even separate any details from them to share.

[eddiew]

Surprisingly good...

I learned by watching MechJeb, tbh. From orbit, adjust your trajectory so it will pass right over the target, allowing a bit of wiggle room for the Mun to rotate underneath you. Burn retrograde until you would hit the ground maybe 20km beyond your target. From here, coast until you see your target, and be prepared to make additional braking burns by-eye. Keep a check on your vertical velocity, and burn directly upwards if it's getting dangerous. You don't want to arrive 1km above your target travelling 500m/s downwards.

I'm not as efficient as MJ is, in that I need more corrections and spend more time descending vertically, but after a while you do get to be a good judge of where you're coming down. Don't be afraid to tweak your descent, even in the last couple of km; you can use SAS radial out to point the ship straight up while you fine tune which way you're drifting  

[The Space Dino]

1 hour ago, Brikoleur said:

Skycrane deliveries are pretty easy: you just use a docking port or stack separator with decoupling force set to zero (or near as) between the surface module and the skycrane module, and put radial engines on the skycrane, if necessary angling them out a bit so the jets don't melt the payload (Twitches are good, and Spiders work for very low-grav bodies/very light payloads; Thuds work for very heavy payloads). You land them the usual way, decouple after landing, and fly off with the skycrane.

It's even possible to deliver a stack of modules this way – decouple the bottom-most one first, then take off with the second, deliver that, repeat until out of modules – although if the stack gets very tall things get hairy as it's likely to tip over.

In theory it ought to be possible to recover modules with a skycrane as well, but that would require precision piloting that's beyond my abilities at least, as you'd have to land on the target module's docking port to a precision of tens of centimeters.

Thanks for the good tips Birkoleur!

But how do you squeeze one into a rocket?

1 hour ago, pandaman said:

I think anyone who's ever tried it will have an epic fail or two.  I've had so many I can't even separate any details from them to share.

Oh lol...

21 minutes ago, eddiew said:

Surprisingly good...

I learned by watching MechJeb, tbh. From orbit, adjust your trajectory so it will pass right over the target, allowing a bit of wiggle room for the Mun to rotate underneath you. Burn retrograde until you would hit the ground maybe 20km beyond your target. From here, coast until you see your target, and be prepared to make additional braking burns by-eye. Keep a check on your vertical velocity, and burn directly upwards if it's getting dangerous. You don't want to arrive 1km above your target travelling 500m/s downwards.

I'm not as efficient as MJ is, in that I need more corrections and spend more time descending vertically, but after a while you do get to be a good judge of where you're coming down. Don't be afraid to tweak your descent, even in the last couple of km; you can use SAS radial out to point the ship straight up while you fine tune which way you're drifting  

That bunny works surprisingly well.... for a weird design and compact lander!

Well done eddiew!

[Guest]

Just now, The Space Dino said:

But how do you squeeze one into a rocket?

Normally you don't need to squeeze, only the radial engines protrude to the sides a bit, it's otherwise the same form factor as the payload. You can make them in any size from 1.25 up. I'll post pictures later if you like.

[The Space Dino]

1 minute ago, Brikoleur said:

Normally you don't need to squeeze, only the radial engines protrude to the sides a bit, it's otherwise the same form factor as the payload. You can make them in any size from 1.25 up. I'll post pictures later if you like.

Please do post pictures !

As a not-very-good KSP player, pictures will definitely help me !

[pandaman]

Ooohh, a couple of mine spring to mind actually...

My first 'successful' Mun landing got stranded due to running out of fuel a few meters above ground, landing a bit heavy and falling over. This resulted in Jeb planting a flag and waiting until I could work out how to rescue him.  After several crashes of rescue ships I got a 'close' landing (16Km away kind of close), which also fell over, but had lots of fuel.  He had to walk/jetpack to it then had to take off horizontally by running up a crater side.

Another, on a mission to plant a flag at the Mohole (North pole on Moho) the fuel ran out on final descent (again), the accuracy was pretty good, but Enbart had to bale out to land using jetpack and his trajectory took him down into the hole, which actually saved him as he wouldn't have slowed down enough before reaching 'normal' ground level, but he then got trapped in the bottom when he clipped through the terrain mesh, so I had to terminate the poor guy in the tracking station.

[W. Kerman]

I have landed things 500 meters apart, but that was on a northern Minimus base, and I got within 4 km of something on Duna.

Not trying the Mun anytime soon.

[eddiew]

1 hour ago, The Space Dino said:

That bunny works surprisingly well.... for a weird design and compact lander!

Near Future is amazing (a little OP, but then, it's far beyond the end of the stock tech tree anyway). The Munbunny had nearly 8km/s of delta-v in it, which makes it ideal for pinpoint landings since you never have to worry about making course adjustments  

[Sharpy]

Navball minigame similar to the one with docking. Yesterday I had to abort the landing, hover and correct (spoil) the Jumperlab landing place not to crash into the Surveyor which I had selected as the target (landed next to the Minmus base) and aimed my landing too precisely.

 

The Navball game is similar to docking approach, but played only around the Retrograde direction.

First, lose most of horizontal speed while still in orbit - approach in a steep, nearly vertical trajectory. Set "Speed" display to "Surface".

During descent aim your ship roughly retrograde (skywards) and note the location of the anti-target marker.

Tilt ship in the opposite direction from retrograde than the anti-target. If you have a lot of speed to lose, tilt it just a little. If you have little, tilt more. Regardless, the farther the Retrograde marker from target, the more you should tilt. - up to the horizon line.

Throttle up some.

The retrograde marker will begin crawling away from your center of navball (the -v- mark) - and towards the anti-target marker since it's in the opposite direction. You can throttle up if you see it moves slowly. Get it as close as you like and shut down the throttle.
 Repeat once it strays again.

You will notice, that as you are accelerating towards the surface, the retrograde marker crawls towards the 90 degrees dot (center of blue half of navball) so you may want to push it "below" the anti-target marker instead of right onto it, so that it stays "on target" longer.

 

[Guest]

@The Space Dino These aren't actual functional spacecraft, I built them to demonstrate the principle. The payload/skycrane proportions are also fairly representative of the real thing, i.e. the payload is usually a little wider even. I did test them on the launchpad -- delivered the payload to the nearby field and flew off (to my doom). I.e. they have way too much thrust to weight for most real missions as they can fly even under Kerbin gravity. The rover sizes are also similar to what I really use; the light one would be a prospector/science rover, the heavier one could be a mining station module.

Both are Twitch-powered: the very light one has two, the heavier one has six, and can only fly in Kerbin gravity with less than full fuel load. Under Mun gravity the heavier one could make do with three Twitches easily (including a full fuel load), while the light one could use Spiders -- four would certainly get the job done but it might work even with only two. Other than that, there's nothing there you can't see; the stack separator has the ejection force set to zero so your crane won't fly off when it separates.

[klesh]

Indeed, set the previously landed rover as a target and pretty much just approach it like you were rendezvousing and docking to it.  You'll be landing right next to things in notime. 

http://steamcommunity.com/sharedfiles/filedetails/?id=536432659

http://steamcommunity.com/sharedfiles/filedetails/?id=691496516

http://steamcommunity.com/sharedfiles/filedetails/?id=691497129

http://steamcommunity.com/sharedfiles/filedetails/?id=691497782

Edited September 23, 2016 by klesh
Added some images.

[suicidejunkie]

On Minmus, I land my puff powered shuttle on the X-Pad's docking port jr.  (9 struts arranged in an X for stability with the center one pointing upwards for mounting widgets)  That does involve coming to a hover within 10m, although the descent is fairly slow for the last km.

 

On Mun, with decent TWR, during the day, I usually overshoot a little and then come back during the final descent.

I haven't had many targets to aim for lately, but I landed within 100m of the Memorial after I spotted it with lots of fuel left in an SUV grade mining hopper.

The second targeted landing was my spaceplane; this was at night, with low TWR, and running on fumes.  It came down 4km from munbase alpha, on the slope of a surprise crater with 17 units of fuel left.  I spent most of that climbing the crater only to find another crater on the other side.  They're now waiting for sunrise to get the lay of the land; if I can get to flat terrain with the 7 units remaining, then the RCS will be enough to taxi to the base for refuelling.

 

To me, the landings are four phases.

1) Orbital Phase

1a) 180 degrees out, lower PE to above the target site, as low as is safe while maintaining enough clearance for mountains and an eye for the body's rotation.

1b) 90 degrees out, do any gross inclination adjustments required to visit not-quite-equatorial sites. 

1c) Set a maneuver node for braking to a dead stop when directly above the target site.  This is only a rough guide and not to be actually followed.  Note: Do this before switching scenes.  This gives your braking burn time, since your ship's expected acceleration should be stably known after the previous burns.

2) Deorbit Burn

Begin the braking burn with the burn time spread about 3/4 of the time before the node and 1/4 after.  I start out with a horizontal burn to allow the ship to drop past the hills I was avoiding, and then transition to retrograde when vertical speed is reasonably high for a prompt descent.  Angle gently to fine-tune push the surface retrograde to the same heading as (but still well below) anti-target.

3) Flight

When horizontal speed is down to roughly 1/3rd of orbital speed, start angling more vertically to keep retrograde below anti-target as you approach.  Lots of seat edges, pants and eyeballs will be used here.  Near the end, divert as necessary to find a clear landing site.

4) Stop, drop and don't roll.

Once you're almost over the site and nice and low, hold retrograde and burn at close to 1 TWR, nudging the throttle up and down to control your speed.  Horizontal speed will be killed naturally.  Touch down and apply the parking checklist for your vehicle.

[IronCretin]

25 meters, with a lander and a science package. I landed a rescue lander a similar distance away.

[gogozerg]

With practice and technique (see @suicidejunkie's post), you'll see it's not difficult to land inside a <200 meters circle around your target on Mün.

 

For a higher precision, things are getting tough.

On Minmus, it's much much more easy. You can comfortably land vertically close to your base (about 20 meters, in order to plug a KAS pipe and refuel). Well, it depends a lot on your ship too.

On Mün, not doing things right can cost a lot of dV. Frequent and reliable operations (I hate F5) require wheels for the last meters.

 

 

[problemecium]

I used to worry about this. Then I got gud skrub er, ahem. Well technically I did get better, but more relevantly, I recently took to a habit of packing the whole base in a tube and landing it all at once. Check my signature for some (old) examples.

[The Space Dino]

11 hours ago, Brikoleur said:

@The Space Dino These aren't actual functional spacecraft, I built them to demonstrate the principle. The payload/skycrane proportions are also fairly representative of the real thing, i.e. the payload is usually a little wider even. I did test them on the launchpad -- delivered the payload to the nearby field and flew off (to my doom). I.e. they have way too much thrust to weight for most real missions as they can fly even under Kerbin gravity. The rover sizes are also similar to what I really use; the light one would be a prospector/science rover, the heavier one could be a mining station module.

Both are Twitch-powered: the very light one has two, the heavier one has six, and can only fly in Kerbin gravity with less than full fuel load. Under Mun gravity the heavier one could make do with three Twitches easily (including a full fuel load), while the light one could use Spiders -- four would certainly get the job done but it might work even with only two. Other than that, there's nothing there you can't see; the stack separator has the ejection force set to zero so your crane won't fly off when it separates.

Thanks for the helpful tips @Brikoleur!