Satellite Constellation Phase Angles Getting Altered Over Time

[neilcomara]

Hi guys, I've been setting up evenly spaced triangle orbit satellites at 120 degrees for my network and can get them at exactly at 120  from each other without any deviation.  But every six months-year the phase angles either rise or drop by .01 degrees and decades later my little triangles are all messed up.  I know it's because my apoapsis and periapsis are just barely not matching but damn it's hard to get the apoapsis and periapsis that precise.  Has anybody else had this issue?  

[SpeedDaemon]

There's no way to get a satellite constellation to stay put using in-game methods that I know of. Maybe a mod I haven't found yet...

Here's a post of mine from last year on how I handle this:

Quote

Well, there is, sort of... The way I've handled this in the past is to build my satellites as if there was some kind of station-keeping facility (usually just including RCS thrusters, and 2 or so units of monopropellant per year of planned lifetime).

Once they're launched into the proper orbits, edit the save file to make all of the satellites in the constellation have exactly the same orbital parameters, with no digits after the decimals, and the proper phase separation. They will stay perfectly lined up as long as they stay on rails (you never fly them, or get close enough to load them into physics). I don't really consider this to be cheating, given there's no other way to do it that I know of, and I'm spending the extra mass/funds to outfit them for station keeping. Same technique might work for placing craft at L4/L5.

Once their station-keeping fuel is "used up", you can send a "refueling" mission, or replace them, or whatever.

 

[neilcomara]

Awesome I'll have to try that, thanks for the great advice!

[bewing]

It's not the Ap and Pe that you have to match. It's the orbital period. And with some carefully throttled thrust from some low-power engines, you can get the period difference down to a fraction of a second. Which won't visibly drift.

[Snark]

7 hours ago, neilcomara said:

Has anybody else had this issue?

My own solution to this problem is not to try to solve it at all.  I just make a point to always have a relay antenna on everything I launch (the HG-5 works great for low-orbit constellations, fits anywhere, and doesn't weigh or cost much).  As my career progresses, I get a fair amount of hardware in orbit.  None of them even try to match each other's period.  Pretty quickly, it gets to the point where I have basically guaranteed coverage, with zero effort on my part.

[LN400]

Bewing is right. If you use a mod that gives you detailed flight info, like KER, and some auto throttle capable mod like RT then it is quite possible to have the orbital period match to tenths of a millisecond.

[DrLicor]

As @bewing said, it's the orbital period. For a kerbin geostationary orbit, make sure all satelites have a orbital period that match the Sidereal rotation period: 
Which is 21549.425 seconds, or 5h, 59m and 9.425s.
You can use KER to readout the period of the satelite. Make sure every satelite has the same period on 3 decimals precise. It doesn't matter if the periode of all the satelites are a bit of, as long as they are all the same off. 
You can use the RCS thrusters to get these precise number right, turn them down to 1% thrust (and enable precise mode hitting Caps-Lock), and gentle tap the shift or control key while in docking mode. When the right number has been reached, disable RCS and SAS to prevent movement and orbital drift. Do this for all the satelites, and you're good to go for a few 100 years.  

 

[neilcomara]

Awesome, thanks guys!