May be a silly question about AP PE

[chenzo]

Hi all. Ive noticed that sometimes when i play my ap and pe will bounce wildly. Just wondering why as im useing the same ship every time i fly, yet it only happens some times

Edited March 4, 2014 by chenzo

[wasmic]

Due to KSP precision errors, your ap and pe slowly jump around a little. If the difference between the ap and pe is very small, it will make them bounce around a lot or even switch places for almost circular orbits. For large enough ships, turning the ship around is enough to switch the ap and pe if the difference between them is less than a few hundred meters.

[Starstrider42]

As far as I can tell this is a problem with limited computational accuracy in the code that calculates spacecraft trajectories. It should only happen when

1. You're in real-time mode (the orbit should freeze in time warp)

2. Either your orbit is very close to circular, or the periapsis passes very close to the planet. In the first case, a small uncertainty in trajectory turns into a big uncertainty in the periapsis location because all points along the orbit are at almost the periapsis (or apoapsis) height anyway. In the second case, I assume it has something to do with close approaches requiring smaller simulation steps, but I don't know the details of how the engine handles that (e.g., an adaptive step size).

[Nepos]

I notice that when mine is jumping and I want to set a nav node I just put the game in 5x time warp and the bouncing stop. this allows me to set the node without all the jumping. Not sure if there is a downside to this other than the time speed increase.

[Red Iron Crown]

I thought it was tidal forces that cause this, and it's more pronounced in certain orientations of the craft than others.

[chenzo]

changing speed has helped out loads with my node planing . Isnt a major problem really.. just alittle annoying

[Kasuha]

It is caused mainly by the fact that the ship is represented by its root part, not by its cener of mass. As the ship is wobbling and twisting in space, the root gets pushed in various directions and its speed in space changes slightly which results in changes to the whole displayed orbit.

This is probably not the whole story but what I know about it so far corresponds to it.