Mun Landing Issue

[Dreamkruiser]

I am attempting to land a base module on the mun, but instead of having rockets on the module I am lowering it by having it attached to a docking ring on the bottom of a descent vehicle. The bottom unit with landing legs is a hitchhiker unit with a crew cabin docked at 90 degrees and has a couple of antennas attached, so I know it's off center, but I thought not enough to be a huge problem. Using RCS and SAS set to retrograde I can descend successfully until about 100 to 200 meters, and less than 50 m/s relative to surface then the craft starts to kick around and it's impossible to get it lined up before hitting the surface. Full throttle prograde does no good because of the imbalance and the outcome is the same. Any idea how to compensate, that hopefully doesn't include removing the attached crew cabin? Maybe I can send another engine out there and attach it to the crew cabin?

[Aegolius13]

You could dock another crew cabin or similar counterweight to the other side if a spot is available.  Could even drop it once on the surface.  

If no port on the other side, you could launch a  counterweight and attach it to the other size using the Klaw.  

If you're amenable to using KAS, you could put some RCS thrusters on the heavy side, as far out from the CoM as possible to provide come compensating torque.

Your extra engine idea sounds feasible, but I'd suggest testing in the VAB to make sure you know where to attach it.  Though you could probably use the throttle limiter slider to fine-tune the balance point as well.

(Also, welcome aboard!)

[HvP]

WELCOME to the KSP forums!

2 hours ago, Dreamkruiser said:

set to retrograde I can descend successfully until about 100 to 200 meters, and less than 50 m/s relative to surface then the craft starts to kick around and it's impossible to get it lined up before hitting the surface.

This actually sounds more like what happens when you slow down so much that you actually start rising upwards again. Pay close attention to your navball. Is it flipping between the prograde and retrograde markers near the surface? If it does then retrograde SAS will usually just default to standard SAS hold, but it has been known to try and flip your craft over to keep pointing retrograde - except that now retrograde is towards the planet's surface. The solution is to either make sure that you watch your velocity (in the window above the navball) and control your throttle carefully so that you don't arrest your descent too early. Or you can switch to normal SAS hold once you slow down enough.

Another possibility is that your engine gimbal is providing most of the control that's keeping you stable until you lower your thrust so much that there isn't enough thrust to compensate for the off center torque. Except for a redesign to center your mass better I doubt you have many options. Instead of another engine you could try sending up a similarly weighted fuel tank with a claw on it and stick it to the other side of your lander as Aegolius13 suggested.

Or, since you are able to get to within 50 m/s of landing then you might have some luck trying to pull off a well executed suicide burn - where you wait until the last possible moment to activate engines at full throttle so that your velocity reaches zero at the exact moment you touch down. Calculating this would be difficult and probably requires several attempts with ample use of quicksave/load.

Without knowing more details about your craft design it would be hard to be more helpful, though.

Edited February 27, 2019 by HvP

[bewing]

Have you scouted the terrain where you are landing? Do you know for a fact that it's extremely level and its exact altitude? Because you aren't going to have any attention to spare for the actual touchdown details.

I've tried tricks like this myself, and it doesn't come out well. The problem is that your off-center mass makes your thrust not be truly retrograde. Which means you build up a horizontal velocity. Which will kill your ship when it touches the ground, unless you're better at flying by hand than I am. Basically, you're going to have to set your SAS to stability mode, and maintain just the right slightly-off-retrograde thrust angle so that it balances out the torque from your crew cabin.

 

[Kryxal]

For this sort of situation (it was actually an asteroid, but it's the same idea) I used 4 nukes (no swivel) with plenty of thrust and adjusted engine thrust levels by hand.  It lets you get FAR closer to true retrograde thrust, such that the reaction wheels can dampen any remaining torque.

[Zhetaan]

13 hours ago, Dreamkruiser said:

Using RCS and SAS set to retrograde I can descend successfully until about 100 to 200 meters, and less than 50 m/s relative to surface then the craft starts to kick around and it's impossible to get it lined up before hitting the surface.

 

10 hours ago, bewing said:

The problem is that your off-center mass makes your thrust not be truly retrograde. Which means you build up a horizontal velocity.

As these and others said, your problem is off-centre thrust.  When you get down to 50 m/s relative, the imbalance is a much larger fraction of the total thrust.  I can think of two ways to fix this:

1. Balance the rocket.  Wheels and RCS aren't good enough for this, because while they will orient your rocket to the correct direction, your main engines will still thrust through something other than your centre of mass.  RCS won't cancel horizontal velocity unless you use translational controls (and won't do it automatically in any case), and reaction wheels can't cancel horizontal velocity.  Therefore, you need either to move the centre of mass to the axis of thrust (by adding massive parts as counterweights as @Aegolius13 said) or move the axis of thrust to the centre of mass (by adjusting the thrust limiters on the engines as @Kryxal said).  One results in fuel waste because of the extra mass and the other results in tricky landings because your rocket, though balanced, will not be level.  Also, unless you're using radial engines or another solution that draws fuel from a central tank, the imbalance will reappear because uneven thrust means uneven fuel use.
2. Learn to fly by hand.

   10 hours ago, bewing said:

   Basically, you're going to have to set your SAS to stability mode, and maintain just the right slightly-off-retrograde thrust angle so that it balances out the torque from your crew cabin.

   And that's really what you have to do.  Knowing that is not the same as knowing how, but it can be done.  Alternatively, you can always build a little excess horizontal velocity into your approach vector and orient your rocket so that the imbalance kills that velocity rather than adds to it, but you will still need to do that by hand.  Don't expect accuracy, and hopefully you have a way to move the module on the ground if it needs to attach to anything already there.

I don't think that the problem is that you're rising again, as @HvP suggests, because if that's happening when your surface velocity is 50 m/s, then it means that you have 50 m/s of horizontal velocity--which means that the problem is still horizontal velocity.  I do agree that engine gimbal is probably helping you until you get to low velocity, but that's because gimbal is a way of moving the axis of thrust.  It's usually better to deal with the problem at the source (by realigning the thrust axis to the centre of mass) than to rely on gimbal or RCS or what-have-you to compensate.

Otherwise, good luck, welcome to the forum, and do please give us an update on how things work out!

Edited February 27, 2019 by Zhetaan

[Dreamkruiser]

Thank you for all the great tips, I've been stewing on it all day. I'll end up trying all of these and might just ditch the cabin and find a way to land it another way, attaching it on the surface is a problem which is why I did it in orbit. My descent ship has 3 engines, does anyone think I could I perhaps change the slider in one or two engines to force more thrust to the off center side?

[Zhetaan]

1 hour ago, Dreamkruiser said:

Thank you for all the great tips, I've been stewing on it all day. I'll end up trying all of these and might just ditch the cabin and find a way to land it another way, attaching it on the surface is a problem which is why I did it in orbit. My descent ship has 3 engines, does anyone think I could I perhaps change the slider in one or two engines to force more thrust to the off center side?

Certainly.  In fact, with three engines, assuming that they're in a triangular configuration instead of a linear one, there is a solution which will perfectly balance the axis of thrust with the centre of mass (within reason; the centre of mass still has to be between the three engines).  Of course, KSP doesn't tell you the information you need to accomplish that balancing once you're in flight, but it is possible.  You may consider duplicating the vessel in the VAB and adjusting the engine thrust while you have the centre of mass and centre of thrust indicators activated.  When you get slider values that point the thrust vector through the centre of mass, simply use those values for the engines of the vessel in flight.

[Kryxal]

Making your adjustments on the fly involves bringing the thrust up with SAS active and seeing which way it ends up trying to turn to counter the imbalance.  That's a great deal easier with a longer burn.

[Dreamkruiser]

Thanks folks for everything on this particular issue. Descending was way easier when I reduced my outwardmost engine to 89%, I was even able to pull off a suicide burn, however after a dozen attempts or so the best case scenario is my fist section laying on it's side with all peripherals attached. No matter how smooth I came down it always fell on a side, but it should be easier to get it upright than land perfect. Thanks again

[Snark]

On 2/27/2019 at 4:16 PM, Dreamkruiser said:

Thanks folks for everything on this particular issue. Descending was way easier when I reduced my outwardmost engine to 89%, I was even able to pull off a suicide burn, however after a dozen attempts or so the best case scenario is my fist section laying on it's side with all peripherals attached. No matter how smooth I came down it always fell on a side, but it should be easier to get it upright than land perfect. Thanks again

Great, glad you got it sorted! 

By the way, one little tip that can sometimes be useful for landings on vacuum worlds:

Normally, most of your descent you'll want to set SAS to hold (with the navball set to surface mode) all the way down.  That does pretty much exactly what you want it to, for an optimially efficient burn with no cosine losses that will gradually get you to a vertical descent as you slow, finally touching down at some speed low enough to not break anything.  That's what I do, and it works great for about 95% of my landing circumstances.

However, it does sometimes run into issues if you ever need to slow down to something close to a hover, for some reason, and/or if your ship is unstable at all.  That's because when your surface velocity gets really low, the marker's position can get unstable and wobble around a lot, which can cause SAS to go nuts and the ship suddenly starts spinning 'round.

One thing you can do, in such circumstances, is this:  when you get slowed enough that you're basically heading straight down relative to the surface, switch SAS to hold instead of .  That may sound nuts to you, if you're used to what these options mean when the navball is in orbital mode.  However, when the navball is in surface mode, the SAS setting means something very different:  it means "point straight up", i.e. directly away from the center of the planet.

That's nice, because as long as you have no remaining horizontal speed, "just hold my nose pointing straight up, regardless of my velocity vector" is exactly what you want, and not subject to any instabilities in the direction of the vector (e.g. it doesn't flip 'round 180 degrees if you accidentally thrust too much and start rising, that sort of thing).

Like I said, I find that it doesn't come up all that often, but it's a handy thing to know about-- just a useful tool in the toolbox for when circumstances warrant.