2013-11-18

Four Zeros Of Orbital Eccentricity

So due to copious spare time, the KOS project is grinding along slowly.

It turns out that the hardest part of the launch sequence is getting a halfway-useful circular orbit.  The math is hard to figure out, and it is difficult to work out a spit-ball method of trying to do it reactively.

So, I cheated.  I abandoned 0.8.4, which had a bunch of exciting bugs, and am now running 0.9.2 instead, which has different exciting bugs, but fewer ones appearing in the currently useful command set.

More to the point, I also found these scripts on the KOS wiki -- a useful, if occasionally misleading, resource -- and with a bunch of tinkering they get the job done.

My changes are not that significant.  I threw away the warp handling because it didn't work for me.  The revised node execution script now tells you that you are in a warp-permitted window, then unwarps you forcibly 60 seconds before the node should be executed.

The net result is this:

Four zeros of orbital eccentricity.
KX-06 is in a 110x109 km orbit.  That's pretty damn good.

The KX-series is the same launch and orbital vehicle as the KT-series.  I called them KX instead since I am experimenting with someone else's code, not testing my own.  But that's ego for you.

As you can see, since the last update, I've been firing off rockets like crazy, with varying results:


...but to be fair I should mention that some of those were experiments with geo-synchronous (kerbo-synchronous?) orbits that didn't pan out for whatever reason.  KSP reports 19 probes in orbit at the moment.  And there have been a ton of launches that didn't pan out.  Over all, what with reversions and all, this has probably been in excess of 60 launches.

But it's all for science, right?

I've also been strict about by debris.  As of right now there are only three sets of orbital debris -- one for KX-06, which is still in its launch ballistic track,  and two tied to the kerbo-synchronous attempts which will take several days to return back to Kerbin.  End result is that I've got a lot of stuff in orbit, but nothing that is junk.  Well OK, nothing that is all junk, anyways.

I feel bad that I cheated, that I wasn't smart enough to figure this out.  Part of my problem is that I didn't know what KOS exports in terms of usable values, and the ones I was playing with for the KT-series just didn't have the information in them that I needed.  In reading through the code I've learned enough that I could probably code a reactive circularization module -- but frankly doing it "correctly" with calculated values is a much better way to do it.  And I've stared at enough wikipedia pages on eccentricity and semi-major-axis et al that I think I understand the math in the code.  I might re-code the module myself to ensure that I really do understand it, but I'm not sure that's necessary.

So what's next?  The original plan was to use the 100x100 orbit as a "staging" orbit and then calculate ways to transfer probes to more useful altitudes.  I was going to write a series of programs to automatically put satellites into kerbo-synchronous orbit with particular destination slots, just because.   But that was before circularization started to kick my ass.  So for now I don't really know what I'll do.

But frankly there's so little reliable spare time these days that I probably have quite a while before it becomes an issue.