Incorrect ascent guidance circularization burn
Kronos9326 opened this issue ยท 4 comments
Hi,
With Mechjeb 2.0 many of my rocket designs fail to achieve as accurate of an orbit during the circularization burn - typically with greater than 5 or 10km differences in Periapsis and Apoapsis (normally 70km x 80km or so if trying to reach 75km x 75km). I removed 2.0 and went back to 1.9.8 and the same design flew perfectly to orbit (75km x 75km).
I've since narrowed it down to having a stage separator on the upper stage, one that separates the payload and the stage itself. If I remove all stage separators from the upper stage I get nearly a perfect orbit insertion ( 74.5km x 75.5km or so) but if I add it back without changing anything else I get the previous inaccurate orbit. It's not a huge problem but it is odd to me. I haven't been able to find anyone else having a similar problem. I'm sure there's probably something obvious I'm missing but I can't figure it out.
I've also tried enabling and disabling auto-staging, corrective steering, and adjusted the ascent profile and final orbit that's desired to no effect.
I've tried Mechjeb 2.0.8, and the latest dev builds, as well as completely deleting KSP and reinstalling fresh with just Mechjeb 2.0.8 added and the same effect occurs.
Thanks; if you could provide a .craft file and a list of all your ascent settings, particularly the turn shape parameters, that would help us reproduce the problem here.
After more testing with various designs it seems a stage separator doesn't necessarily cause this problem for all designs. Below I've two different rockets with the only difference being the upper stage engine. With the ascent guidance settings being 75km orbit, 0 inclination, 10km gravity turn, 40% turn, 75km turn end, at 0 degrees and limit to terminal velocity, auto-staging enabled, and corrective steering enabled. Once "coasting to edge of atmosphere" is reached I jettison the first stage.
With MJT4.craft it'll on average achieve an orbit of 80km x 70km instead of 75km x 75km. But with MJT4_2.craft it'll achieve around 76km x 74km or better. But using MechJeb 1.9.8. they both reach the same, perfect orbit. With 1.9.8 I notice it burns at apoapsis and angles the nose up properly to stay at apoapsis during the burn, but MechJeb 2.0.8 just blindly executes the burn it planned prior without adjusting the vector to achieve the exact orbit.
I don't have KSP open at the moment so I'm not sure about the details of your craft, but from my own experience I can tell you it helps to have a shorter circularization burn. You can do this by ending the gravity turn at a lower altitude, so you burn more horizontally, raising your periapsis higher before your apoapsis reaches the target orbit altitude. Try experimenting with the turn end altitude, try 50km, 60km, etc and see if it improves.