Using RCS thrusters to fine tune orbital insertion (Pe) after initial burn

[JoeSchmuckatelli]

Does anyone else do this?  I just discovered it. 

 

I had to do another Minimus run to gain some cash for other missions - and goofed my initial burn.  The correction burns I tried - kept over-correcting - either giving me too high a Pe or having me miss the encounter entirely.  But rather than reverting to a previous save - just for kicks I thought I'd see what the RCS thrusters would do - and they gave me an almost perfect insertion Pe.

 

I'm certain I'm not the only one who does it - but does anyone do this regularly?  It seems to work very well.  Especially where you've one ship with a lot of Dv that gives you fits trying to fine-tune an insert.

[ForScience6686]

That is one method.  Or you could limit the thrust in the engine right click menu, if you don't have rcs.

[ElWanderer]

RCS usually has the advantage that you can slightly alter your orbit in any of six directions without having to change the way you're pointing. Not having to turn around if you overshoot is very handy.

[JoeSchmuckatelli]

19 minutes ago, ForScience6686 said:

That is one method.  Or you could limit the thrust in the engine right click menu, if you don't have rcs.

Learn something new every day.  Thanks - that will come in handy!

[Snark]

Nothing wrong with using it as a (very) fine-tuning technique.  In general, it's a bad idea compared with main engines, because RCS has a fairly crappy I_sp.  But if you're only using it for very small nudges, then the convenience factor could outweigh that.

I usually deal with orbital fine-tuning by just lowering the thrust limiter if needed, then pointing the way I want.  I deal with overshooting problems by just being careful and not overshooting. 

[rodentgun]

How is the best burn? I try to split my burn across the navigation timer 0 point. But near the end of the burn the blue target drifts and I somehow still end up overdhooting a wee bit. Is it how it is?

On 28. dec. 2015 at 3:48 PM, Randazzo said:

I suspect you'll see a lot of references to the tyranny of the rocket equation pop up in here. Basically, you start get diminishing returns rather quickly in terms of adding more fuel/thrust. You need a little more dV so you add more fuel only to discover your TWR is too low, so you add more thrust, only to discover you need more fuel, so on and so forth. This is why rockets use stages, and specific to your situation, why the Apollo lander had two stages. You probably won't need to go to that extreme just to land on and re-ascend from the Mun, but if you want to return with the same vehicle, a delivery stage would help.

I've probably rambled a bit farther than necessary, but the answer to

is: Stages

 

[Randazzo]

@pistolhamster Accidental quote?

Orbits are never going to be perfect in-game due to floating point errors, and nailing the burns perfectly is nigh on impossible for a human being. You can get better with practice but you'll never be absolutely perfect, unless you're a computer, and even those can have a hard time.

You want to burn for precisely the same amount of time before the node as after it, for the "best burn". Really the best burn would be instant acceleration on the node, but that's impossible.

[Snark]

34 minutes ago, pistolhamster said:

But near the end of the burn the blue target drifts and I somehow still end up overdhooting a wee bit. Is it how it is?

You're doing fine splitting your burn across the node.  That's not the issue behind the "drift".

The "drift" is a natural consequence of getting close to the end of your burn; it's just a magnification of any minute errors present from the start, and easily corrected for.

Think of it like playing golf:  You drive off the tee, and the hole is several hundred meters away.  It's possible to get a hole-in-one, but very rare.  More likely you get it fairly close to the hole, and when you get close up, you need to do a smaller correction shot to sink it.  Since the initial drive is so long, you can shoot it with even a tiny fraction of a degree error and it will still miss the hole, even though it looks perfect to you as it leaves the tee.

Imagine that there's a camera on the golf ball that's pointing straight in the direction of where the hole is.  (That's the equivalent of the maneuver marker on the navball).  When you fire the ball, with only a tiny fraction of a degree error, then everything looks perfect, right?  The camera on the ball is pointed "straight" at the hole (actually it's off by that tiny fraction, but it's too invisibly small to tell.)  But as the ball gets closer and closer to the hole, the error becomes more and more obvious.  The camera on the ball, still looking in the original direction, will see the hole "drift" off to the side..

So the way to fix this is to expect it (it will always happen, always) and correct for it as it occurs.  That is, start your burn pointed straight at the maneuver node, as best you can.  Watch it as you burn.  As the remaining dV gets smaller and smaller, you'll see the maneuver node marker start to drift away from the central cross-hairs.  When that happens, just rotate your rocket a bit to re-center the node on the navball.

The smaller the dV remaining, the faster the node will drift.  To pinpoint it, what I usually do is don't try to do the whole burn.  Kill the thrust when there's still a few m/s remaining on the burn.  Then carefully align the node in the center of the navball and do a few brief, controlled spurts of low thrust to finish it off.  Using this technique, it's pretty easy to get an "exact" burn (i.e. error of less than 0.1 m/s) every time, with no overshooting.

[FancyMouse]

When I have tiny stuff (1t payload or so) that needs docking, and I don't want to do a pure-engine dock (I can do it, but I don't want to do it on a regular basis), I'll stick RCS ports and use RCS as main engine - not just for circularizing, but with several hundreds of m/s it can even do things like returning from Mun/Minmus station.

Really for tiny things I don't think it's worth adding LFO engine (along with LFO fuel) when I need RCS anyway. The only bad side is that executing a burn becomes harder, but I'll take it.

[Geschosskopf]

On ‎1‎/‎15‎/‎2016 at 1:11 PM, JoeSchmuckatelli said:

Does anyone else do this?  I just discovered it. 

I used to do this quite a lot, but that was before we got asteroids (0.23.5 I think that was).  When asteroids came along, Squad made some nice changes to the maneuver node system, such as adding mouse wheel input, the ability to have nodes happen several orbits into the future, etc.  However, at the same time, they introduced some bugs in the way your future path is drawn, making it somewhat wobbly.  The wobbles are especially evident when you rotate the ship.  Have the ship pointing prograde and observe your Pe.  Then rotate the ship to point radial in and note how much your Pe changes, which is absurd.  You didn't change the velocity, you just changed the orientation.

The change in trajectory just from rotating the ship is much worse if that ship has RCS installed (but not activated), and is quite significant if you actually use the RCS to reorient the ship.  Even if you used RCS BuildAid so your RCS thrust is perfectly balanced.  Suffice to say that the changes in trajectory due to reorientation with RCS installed but inactive is greater than the tweak you'd made using the RCS to begin with.  And then, if you reorient the ship again after using RCS, the change you get from that undoes all your deliberate tweaking. 

So bottom line is, to me RCS has been unusable for trajectory tweaking since 0.23.5.  The inaccuracies in the trajectory-drawing system are greater than the changes you want to make with the RCS.  I find this very annoying and it's one of the things I most want to see fixed in 1.1.

[MagicFireCaster]

yup i also do it, is a very nice way to get small corrections on your orbit too