help setting up resonant orbits for comnet on mun

[putnamto]

using this tool https://meyerweb.com/eric/ksp/resonant-orbits/ i was able to set up a nearly perfect triangular orbit around kerbin, yet when trying to do the exact same thing on the mun i get different results, im not good at explaining things.

im trying to set up a triangular constelation evenly spaced at 300,000 m, i do my burns according to that tool yet my relays end up in weird orbits, like, instead of it coming out to an equalateral triangle i keep getting an isosceles.

a picture to better explain myself, anybody know what im doing wrong? is the tool flawed? my first instinct says that my orbit is to low, but if i was higher up i would have the same results now that i think about it.

Edited December 6, 2019 by putnamto

[VoidSquid]

I followed this tutorial and got my relay network just perfect:

Making sure the orbital period is exactly the same for all satellites is quite important to have a stable configuration, the individual altitudes can vary a bit, doesn't matter.

[putnamto]

holy fricking math batman

[VoidSquid]

Well, if you want to have a perfect relay sat network, that's the way to go  

Save for the math, what I do step by step:

1. Decide what orbit altitude you want to go for, and if it should be a polar or equatorial orbit. Let's for example say, you want a 500km triangular polar orbit constellation around Mun.

2. Build your rocket with three deployable relay antenna modules, which each should have a probe core, a battery, solar panels, a small fuel tank, and a small engine. Launch it, and bring it into a 500km polar orbit around Mun.

3. Check the resonant orbit values at https://meyerweb.com/eric/ksp/resonant-orbits/ , the important data here are the orbital period of 4h:0m:18.1s, and the altitude of the apoapsis of that resonant orbit which is 795.979km.

4. Deploy your first antenna module, immediately go prograde with your main rocket and raise the apoapsis to 795.979km. I strongly suggest you orient your rocket to normal before deploying the antenna module to avoid potential collisions when raising the apoapsis after deployment.

5. Switch to the antenna module, enable the engine, and check the orbital period, which should be close to 4h:0m:18.1s. If the orbital period is longer, go retrograde and burn slightly to adjust, if the orbital period is shorter, go prograde and burn slightly to adjust. What is important here is to achieve the exact orbital period as close as possible, the actual values for apoapsis and periapsis are not that important (if you, for example, end up with a 505x492km orbit, that's perfectly fine).

6. Switch to your main rocket and circularize at the next periapsis (which is 500km, of course).

7. Continue with your 2nd antenna, same procedure as point 4 to 6

8. Continue with your 3rd antenna, same procedure as point 4 to 5

9. Enjoy your triangular relay setup  

 

Btw, another useful link I followed:

https://wiki.kerbalspaceprogram.com/wiki/Tutorial:Setting_up_a_CommNet_system

 

Edited December 6, 2019 by VoidSquid

[putnamto]

i didnt know about orbital periods and their importance, the stock game doesnt tell you that info as far as i know, KER does though.

thanks guys

[VoidSquid]

Stock game does since 1.7, iirc, check the display in the lower left corner, there are a couple of buttons to change between different data to display. But if you're already using KER, you're more than fine, have all data at your fingertips.  

6 minutes ago, putnamto said:

thanks guys

You're welcome, happy to help.  

Edited December 6, 2019 by VoidSquid

[putnamto]

well the one on the mun didnt turn out so well, going to scrap it and try again, but with my newfound knowledge i made this.

HAIL SATAN!

[VoidSquid]

Looks good, well done  

[putnamto]

5 minutes ago, VoidSquid said:

Looks good, well done  

in the last pic, their orbital periods are off by +-5 minutes, but that shouldnt be cause for concern should it?

and something i didnt notice is that the meyer resonant orbit calculator does days in a 24 hour format, and kerbins days are only 6 hours, i sat here scratching my head for a few minutes on that one.

Edited December 6, 2019 by putnamto

[AHHans]

15 minutes ago, putnamto said:

in the last pic, their orbital periods are off by +-5 minutes, but that shouldnt be cause for concern should it?

Well, what fraction of the orbital period are those +- 5 min? How many orbits does it take to completely mess up the formation?

In other words: I'd change the orbits so that they have the same orbital period down to as good as possible (e.g. one second?). (It doesn't matter so much if they have this period with identical AP and PE: as long as the period is identical they'll keep their relative positions.) You may want to lower the thrust limit of the engines on the satellites to do that.

Edited December 6, 2019 by AHHans

[VoidSquid]

18 minutes ago, putnamto said:

in the last pic, their orbital periods are off by +-5 minutes, but that shouldnt be cause for concern should it?

Well, think about it: let's assume you play 10 years Kerbal time. The orbital period from my example is 4h:0m:18.1s. Let's assume that satellite's orbit is off by 5 seconds.

10 Years = 4260 days = 25560 hours = 6390 Orbits. So, after 10 years that one satellite's orbit would be off by 31950 seconds, or 8.8 hours or more than 2 whole orbits... not good.

I try to be meticulously precise up to the very last digit with the orbital periods, with the result that even after more than 30 years after deployment all my relay sat constellation still look like proper equilateral triangles.

Edited December 6, 2019 by VoidSquid

[putnamto]

Just now, AHHans said:

Well, what fraction of the orbital period are those +- 5 min? How many orbits does it take to completely mess up the formation?

In other words: I'd change the orbits so that they have the same orbital period down to at least one second. (It doesn't matter so much if they have this period with identical AP and PE: as long as the period is identical they'll keep their relative positions.)

the orbital period on three of them is 34 days and 5 mins, on one its 34 days and 8 mins and on another its 34 days and 3 mins, i dont have time to continue playing the game, but ill see what i can do later.

[AHHans]

6 minutes ago, putnamto said:

the orbital period on three of them is 34 days and 5 mins, on one its 34 days and 8 mins and on another its 34 days and 3 mins

Ah, O.K. That means it'll take them about 20 years to move one tenth out of (relative-)position. I would still try to correct that, but if that's fine for you then just leave it.

Edited December 6, 2019 by AHHans
fixed typo

[putnamto]

Just now, AHHans said:

Ah, O.K. That means it'll take them about 10 years to move one tenth out of (relative-)position. I would still try to correct that, but if that's fine for you then just leave it.

i am fine with that, but i have to ask, how did you come up with that? what was the math involved? im horrible at math, and i use alot of mods to get past that, but im curious and would like to learn.

and im also a little dumb. i would ask you to draw me a picture if you could.

[AHHans]

Just now, putnamto said:

i am fine with that, but i have to ask, how did you come up with that?

Let's see:

• 34 days and 8 mins and 34 days and 3 mins is a maximum difference of 5 min
• 34 days and 5 mins are:  34 (days) * 6 (hours/day) * 60 (min/hour) + 5 (min) = 12245 min for one orbit
• deviation per orbit: 5 min / 12245 min = ca. 0.00041
• one tenth deviation in: 0.1 / 0.00041 = 243.9 orbits
• in time units: 243.9 (orbits) = 243.9 (orbits) * 34 (days/orbit) = 8292.6 (days) / ca. 400 (days/year) = ca. 20 years

Yes, that last line is a rough estimate. A Kerbin year isn't exactly 400 days long, your orbits aren't exactly 34 days long, yada, yada, yada. Who cares?

15 minutes ago, putnamto said:

i would ask you to draw me a picture if you could.

I probably could, but I'm not going to!

[VoidSquid]

21 minutes ago, putnamto said:

i am fine with that, but i have to ask, how did you come up with that? what was the math involved?

See my example above or that from @AHHans he posted a few minutes ago.  

Long story short: be as precise as possible, and the shorter the orbital period is, the more important becomes precision. It costs you only a few minutes, which is way less time than you'll need to rebuild a constellation every few Kerbin years.

Edited December 6, 2019 by VoidSquid

[putnamto]

40 minutes ago, VoidSquid said:

See my example above or that from @AHHans he posted a few minutes ago.  

Long story short: be as precise as possible, and the shorter the orbital period is, the more important becomes precision. It costs you only a few minutes, which is way less time than you'll need to rebuild a constellation every few Kerbin years.

thats for sure, i did get my mun constelation looking good, but it turned into a little wile later.

[VoidSquid]

5 hours ago, putnamto said:

draw me a picture if you could.

I'm not any good at drawing as well, you'd end up with a drawing similar to this, best case (found this one on the NASA website https://www.jpl.nasa.gov/nmp/ds1/img/top.jpg).

Edited December 6, 2019 by VoidSquid

[linuxgurugamer]

4 hours ago, putnamto said:

using this tool https://meyerweb.com/eric/ksp/resonant-orbits/ i was able to set up a nearly perfect triangular orbit around kerbin, yet when trying to do the exact same thing on the mun i get different results, im not good at explaining things.

im trying to set up a triangular constelation evenly spaced at 300,000 m, i do my burns according to that tool yet my relays end up in weird orbits, like, instead of it coming out to an equalateral triangle i keep getting an isosceles.

With the author’s permission and encouragement, I made this a while ago:

 

[Curveball Anders]

7 hours ago, VoidSquid said:

2. Build your rocket with three deployable relay antenna modules, which each should have a probe core, a battery, solar panels, a small fuel tank, and a small engine. Launch it, and bring it into a 500km polar orbit around Mun.

Actually, you only 2 deployable relay modules, if you use the carrier as the third.

[VoidSquid]

1 minute ago, Curveball Anders said:

Actually, you only 2 deployable relay modules, if you use the carrier as the third.

Eeeek  
That hurts my aesthetic feeling! Not symmetric! And ugly!  

[Curveball Anders]

Just now, VoidSquid said:

Eeeek  
That hurts my aesthetic feeling! Not symmetric! And ugly!  

But efficient and cheaper

 

[VoidSquid]

My aesthetics know nothing about costs or efficiency  

Edited December 6, 2019 by VoidSquid