US2 issues

[esskay]

I was running simulations and noticed that logging out of my WinXP user account while the simulation is still active and back in puts US2 in a rather confused state, failing to show the simulation any more and flickering the application, until I start another simulation.  Has anyone else encountered this?

Also, I was trying to learn about Lagrange point behavior and was testing out the Earth-Moon example.  How accurate is US2's simulation for this?  I was expecting L1 to be more stable than what I am seeing -- it almost immediately drifts toward the Moon and whips around it.  Are the Lagrange points really that unstable or am I not understanding how they should behave?

[Dan Dixon]

I was running simulations and noticed that logging out of my WinXP user account while the simulation is still active and back in puts US2 in a rather confused state, failing to show the simulation any more and flickering the application, until I start another simulation.  Has anyone else encountered this?

That doesn't surprise me. This problem might be solvable, but it's not on the radar at the moment.

Also, I was trying to learn about Lagrange point behavior and was testing out the Earth-Moon example.  How accurate is US2's simulation for this?  I was expecting L1 to be more stable than what I am seeing -- it almost immediately drifts toward the Moon and whips around it.  Are the Lagrange points really that unstable or am I not understanding how they should behave?

Good question. I'm not sure how accurate it is. I know that L1 is inherently unstable (like placing a marble on a beach ball vs L4 & L5 which are like placing a marble in a bowl), but the object at L1 does seem to fall out of orbit very quickly.

I spent a couple hours on this today and did fix identified a problem with the calculated L1 point jumping around when you move the Moon around.

1 - Load up the Earth & Moon simulation
2 - Turn on Lagrange points (from the view controls, click the long bar on the far left)
3 - Use the hand tool to move the moon around. Move it toward and away from the Earth to see the L1 point jump around. (it shouldn't jump like that)

A fix for this will be in 2.0.4.

[esskay]

Thank you for looking into these!  I look forward to your next version.

[Dan Dixon]

I talked today with Tony Dunn who made this:
http://www.orbitsimulator.com/gravity/articles/what.html

And created this:
http://www.orbitsimulator.com/formulas/LagrangePointFinder.html

Based on that conversation:
A better analogy for L1 is like balancing a pencil on it's point that can only fall over along one axis (the line between the two bodies).

And the math that I'm using to calculate Lagrange points for perfect orbits, which the Moon & Earth is not. (and I think there may be a bug in the L1 calculations for some values. This bug isn't yet fixed in 2.0.4.

I'm running some new tests now with a 'perfect' Earth Moon system I'm able to get 1.75 rotations of the moon before the L1 point appears to turn in toward the Earth. This 'feels' better.

I'll include these new simulations in 2.0.4.

[Dan Dixon]

We're going to skip the public release of 2.0.4 and release 2.0.5 next week. There are some problems with texture handling that cause crashes for some users.

Anyway... I believe that all auto calculation of Lagrange issues are now fixed.

In Earth-Moon you can get almost 2 full rotations of the Moon before the L1 point visible moves off of it's point (unlike currently where it moves off the point almost immediately).

The Halo at L1 simulation is also far more satisfying with these fixed.

In short:
Lagrange points will work as expected as of the upcoming 2.0.5 release. Thanks again to Tony (of Gravity Simulator) for the help.

[Sanduleak]

Another great announcement! I cant wait for 2.0.5