Launch Spacing for Communications Satellites?

[Exosphere]

I'm playing KSP with RT2, so I'm currently in the process of putting together a kerbostationary communications relay. Rather than launch this relay all at once, I've decided to launch them one-by-one. My planned relay consists of six satellites in a keostationary orbit: three for the main communications relay with one satellite between each of the three main ones to act as a back up to maintain coverage as the satellites drift in their orbits -- I'm trying to get them right at 35,786 kilometers (I'm playing with Real Solar System installed) by using MechJeb to get the periapsis and apoapsis exactly right, but I know there will be some errors, thus, I've decided to add in some redundancy.

Anyways, is there a method to set up my launches so that each of my satellites ends up in the correct geostationary location? I just tried what seemed to be the most intuitive method (launching each of the rockets just under four hours apart, so each one would be launched one-sixth of a Kerbin sidereal day apart), but that didn't seem to work. Satellites one and three seemed to end up in a kerbostationary orbit at the exact same location, as did satellites two and four (each group of two satellites ended up in a section of kerbostationary orbit that definitely was not one-sixth or one-third apart from the other group). So, it appears that the most intuitive method doesn't work. Is there any set period of time to offset each of my launches so that the satellites *will* end up in kerbostationary orbits 60 degrees away from each other?

Edited January 9, 2014 by Exosphere

[Markus Reese]

As apposed to timing, I used a different method to get all my proper spaced satellites. Timing, if your launch angles are different at all, then your spacing can end up way off.

First, Get the one into as close to perfect circle at altitude you want.

Second, launch one at a lower altitude.

Set the first one as your target to burn to. You can use the Target position, and closest position to get the approx 60 degrees apart. This should get you close, and you can do fine changes to get it more exact. Most important part though will be ensuring that they are as exact as possible to eachother in height or will slowly change spacing. You will want some ion propulsion so you can fix spacing as needed.

[Alchemist]

If you launch with exactly the same ascent profile to stationary orbit, you will end in exactly the same place above Kerbin, becouse the surface nd the positions in obit move with the same rate.

One of the options is elliptic orbit with apoapsis at stationary orbit altitude and period 5/6 day. Wait for several orbits, then circularize the satellite's obit. Also this way you can launch several sats at once and then cirularize them one at a time

[Peppe]

For triangles i would not bother with launch windows. Put it an orbit at your apoapsis and just short of circular. Every pass of the apoapsis check your distances and finalize.

Mechjeb is only accurate to the second. I highly recommend VOID, semi major access of 3 sats within 30 centimeters of eachother. Drift stable basically forever...

Don't know if mechjeb does this, but VOID also has a nifty targetting method to let you see distance to a two targets at the same time.

I just shift around the other two until the distances were equal:

Now if i was going to squeeze in another triangle i could just launch one and target each neighbor and move it forward/back in its orbit until it was equidistant to it's neighbors without affect their positioning. To move something forward in an orbit burn retrograde a little to speed it up for it's next orbit. To move something back burn prograde to make a slightly longer orbit for that lap.

[zeppelinmage]

The way I did a 6-ship kerbistationary constellation is this:

1) I launched one and set it at the desired altitude and period.

2) Launched my other 5 ships into circular orbits below the first.

3) I set the original satellite as my Target and used maneuver nodes to "aim" my satellite at its desired location. As I was still new to the game, I used a protractor for this, looking down on Kerbin. >_> Basically, since the target node will show you where the target will be, it's not too difficult to line them up at 60 degree intervals.

4) Circularize when you reach apoapsis with that satellite.

5) Use RCS (or ions) to fine-tune to your desired altitude and period (6h).

That's without mods at all.

Edited January 8, 2014 by zeppelinmage
Fixed URL

[federicoaa]

I like the way scott manley did. Send the 3 satellites on the same ship, put the Ap at geostationary and the Pe so the orbit is 2/3 of kerbin day. Then, for each time you reach Ap, decouple one sat and circularize its orbit.

[Spyritdragon]

I like the way scott manley did. Send the 3 satellites on the same ship, put the Ap at geostationary and the Pe so the orbit is 2/3 of kerbin day. Then, for each time you reach Ap, decouple one sat and circularize its orbit.

This sounds like a very nice way to do it. Easy to understand, too :-)

[Exosphere]

Well, I've decided to just launch the satellites three at a time using a resonant parking orbit, and it seems to be working. I actually got my original idea to launch at four-hour intervals because of a statement on the Figaro mod page that you could achieve the desired spacing between GPS-style orbits by launching at a set interval (in the case of a GPS-style orbit in stock KSP, one hour). I assumed the same would hold true for launching a geostationary comms network, but it appears not to be the case. Does anyone know why that method doesn't work?

[Markus Reese]

Are you meaning launching at set intervals would put proper spacing? In theory it does. If each launch is exactly the same. Manually though, you turn one second earlier or later than the previous flight, you will end up in a different place :-(

[Exosphere]

Well, that's what it said on the Figaro mod page, which I'm about to work on now that I've finally fixed all the bugs with my Remote Tech relay system. Just so I can make sure I get my GPS system up and running, is that the method you'd use? I'm going to launch four satellites at a time (I'll just use a resonant transfer orbit system to get the four satellites in each 55 degree orbit in the proper spacing), but I was going to use launch intervals to launch each set of four satellites, to make sure each set is properly spaced at 60 degree intervals. Is that the best way to do it? Here's the page (and quote) I got the idea of launch intervals from:

http://forum.kerbalspaceprogram.com/threads/24646-0-23-Figaro-Global-Navigation-Satellite-System-Launch-a-Working-Kerbal-GPS-System

The geometry of your constellation doesn't need to be super precise. You just want your satellites to be well distributed around the globe. Real world GPS satellites are in circular orbits with an orbital period of 12 hours (half the Earth's mean solar day). They are located in 6 orbital planes, spaced 60º apart and inclined at 55º from the equator. The Kerbin equivalent orbital period would be 3 hours and 25 seconds (because Kerbin's solar day is 6 hours and 50 seconds long). A 1588 km high circular orbit has about the right orbital period. You can achieve the 60 degree spacing between orbits by launching at 1 hour intervals.

Since I'm using the Real Solar System mod, I'd have to scale up the interval period because the sidereal day is almost four times as long as in stock KSP, but I got the idea that this was the general method you are supposed to use.

[Alchemist]

Well, I've decided to just launch the satellites three at a time using a resonant parking orbit, and it seems to be working. I actually got my original idea to launch at four-hour intervals because of a statement on the Figaro mod page that you could achieve the desired spacing between GPS-style orbits by launching at a set interval (in the case of a GPS-style orbit in stock KSP, one hour). I assumed the same would hold true for launching a geostationary comms network, but it appears not to be the case. Does anyone know why that method doesn't work?

For equatorial orbits angle between satellites launched with interval T will be

360*(T/T_orbit-T/T_Sideral)

Stationary orbit has T_orbit=T_Sideral that gives zero angle (if you use the same launch profile)

And in that description of GPS network the launch time intervals are not about getting the satellites evenly spaced in equatorial orbit, but about geting them into different evenly spaced inclined orbits, and that's completely another question.

Edited January 9, 2014 by Alchemist

[bonkers838]

I don't think launching at intervals will help if you are planning on doing a geostationary orbit. You launch the first one, put it into geostationary orbit. Now it is sitting right above the command center. You wait four hours, launch the second one, but because the first one was in geosync orbit it is still right above the command center. You will end up putting all of them right on top of each other (assuming same launch trajectory). I think this method could work if it wasn't geostationary depending on various factors, but I'm by no mean an expert