Class E Asteroid changed orbit on me

[Maximillian]

I decided Kerbin needed another moon so I went hunting for a class E asteroid to put into LKO. Found a nice candidate in the tracking station and sent a very small pathfinder probe with a Klaw out to measure the mass and estimate the needed DeltaV for the redirect. Built a (5)LV-N powered pusher ship with enough fuel to move the rock by ~100m/s, more than twice the needed course change for a equatorial aerocapture at ~40Km.

After three different burns everything was set, all I had to do was wait for a minor 4m/s burn once I was in Kerbin SOI.

Or so I thought.

I had other things to attend to in the program that were to happen before the SOI change in just under 70days.

Time passes.

Came back to the redirect mission to find that the orbit had shifted radially by 240m/s in just 10 days.

Is this:

1- A glitch in the physics engine?

2- The result of the changed mass of the rock with my ship still attached? (yes it's big, but not compared to the 3KTon rock)

Anyone seen this before?

~Max

Edited September 9, 2014 by Maximillian

[ratchet freak]

where you time warping during the SOI change? Precision errors are common when timewarp is high.

[Maximillian]

where you time warping during the SOI change? Precision errors are common when timewarp is high.

Nope, SOI change was 70 ish days away. I came back to focus on the mission after about 10days game time and found the orbit had changed.

[Geschosskopf]

There's a bug with the maneuver system that makes your trajectory vary based on the rotation of the ship's thrust axis compared to some unknown reference point. IOW, without thrusting at all, just rotating a ship, you can vary both your current trajectory and the delta-v of any maneuver node you currently have active.

The really insidious thing about this bug is that rotation happens by itself over time due to the ship being in a curved orbit. The ship might still be pointing in the same direction as before, but because its direction of motion changes with the curvature of the path, the angle between the ship's axis and its direction of motion changes over time, and the game seems to treat this the same as if you'd rotated the ship.

My guess is that this is what's happened to you. It happens to me all the time, even if I never change focus away from the ship. And not just with asteroids but even just flying between Kerbin and Mun.

[Red Iron Crown]

There's a bug with the maneuver system that makes your trajectory vary based on the rotation of the ship's thrust axis compared to some unknown reference point. IOW, without thrusting at all, just rotating a ship, you can vary both your current trajectory and the delta-v of any maneuver node you currently have active.

The really insidious thing about this bug is that rotation happens by itself over time due to the ship being in a curved orbit. The ship might still be pointing in the same direction as before, but because its direction of motion changes with the curvature of the path, the angle between the ship's axis and its direction of motion changes over time, and the game seems to treat this the same as if you'd rotated the ship.

My guess is that this is what's happened to you. It happens to me all the time, even if I never change focus away from the ship. And not just with asteroids but even just flying between Kerbin and Mun.

Interesting, I haven't experienced that before. Does it only happen when the vessel is off rails?

[Geschosskopf]

Interesting, I haven't experienced that before. Does it only happen when the vessel is off rails?

The situation is a lot better in 0.24 than it was in 0.23.5, when the bug 1st appeared, but it's still there if you look for it. I guess most folks don't notice it because when they're rotating their ships, they're looking at the navball markers instead of their trajectory on the map. To see it in action, mouse over or click on your Ap or Pe so you can see its altitude, then rotate your ship and watch the numbers change. It's especially noticeable if your future path crosses and SOI boundary, in which case watch what happens to your Pe in the new SOI in conics mode 0. And it also affects any nodes you have set out in the future: the delta-v required, the position of the blue marker on the navball, and the resulting future trajectory after the burn, presumably because the node is on your future path already, and that path is changing due to the ship's rotation.

Anyway, it seems to happen whether the ship is on rails or not, and whether you're warping or not. Naturally you see it most when the ship isn't on rails and you're maneuvering, but it appears that on-rails ships suffer from trajectory drift more significant that the expected long-term accumulation of rounding errors. Such as in the OP's case.

The effect isn't crippling, especially for short trips to Mun and Minmus, but it becomes rather bothersome for interplanetary and asteroid-catching missions. Still, the bottom line is that you just need a little extra delta-V for more correction burns than you used to need. The most damaging part is the impact on Remote Tech. There, you require high precision in the placement of satellites, and then rely on them staying where you put them. Sure, you expect to have to tweak them periodically because total perfection is impossible and rounding errors do occur, but you also expect not to have to do this very often or do relatively large corrections. I mean, they're not crossing SOI boundaries and aren't in really huge orbits, so you'd expect them to be pretty well-behaved. The problem with the maneuver system, however, makes precision very hard to obtain, impossible to keep, and the resulting corrections are both frequent and large.

But back to asteroids.... It appears that any wobble in the ship counts as rotation as far as the bug is concerned. I suspect this is because the wobble makes the ship's CoM move relative to whatever physical part of the ship is the reference point for its position in space. Moving the CoM results in moving the whole ship, as shown by the classic example of "thrusting" a rotating ship by pumping fuel back and forth between its ends. And asteroid tugs are extremely wobbly due to the huge mass on at least 1 side (and usually both) of what is just a 1.25m docking port, even if you add struts somehow. So the problem is usually very manifest when maneuvering one, and the residual wobble after orienting the tug can throw the future trajectory way off while you're not looking, before you change focus away from it. You'd think the wobble would go away when you leave the tug on rails, and maybe it does, but it's still going to rotate relative to its direction of motion, which also seems to count as far as the bug is concerned.

[Aethon]

Hmm. I notice this a lot but chalked it up to the inherent lack of precision in determining long course trajectories in space flight. Another thought in reference to this : as a craft is translated about its' center of mass, whichever probe/port the craft is being controlled from ( probably the furthest one from the COM ) has a slightly different trajectory at different times during the maneuver as the craft yaws and pitches, add in wobble ( and SAS anti-wobble correction ) and I can see how these projected trajectories can change.

[Maximillian]

I've marked this Answered. Thank you for the help.