Final stage seperation

[NFunky]

Hey guys, just had a quick question.  Why do all probes/satellites (particularly early ones) seperate from the final stage of their carrier rocket once they reach the proper trajectory/orbit?  And, is do any of those reasons matter in KSP?

I've always done it, but I have no idea why.  I guess just because that's what I've always seen them do.  In early career though, when trying to save every buck per launch, why ever bother?  Those TR-18A stack decouplers are $400 each, which I could spend on another RT-10 instead = more dV/$.

Does everybody do integrated upper stages and I'm just late to the party?  I'm also talking about stock here, but what about RSS/RO?  Does that change the equation a lot?

Thanks!

[Gaarst]

Most real life rocket separate the last stage in GTO and the probe itself does the GEO insertion. Most probes actually have integrated propulsion systems for altitude control or manoeuvring, and upper stages can't always restart when you want or aren't designed to keep functioning for several months, so they ditch upper stages when they did their jobs and rely on their own systems.

In KSP separating the last stage is only useful if you have engines on your probe, otherwise you can just leave it there and save the cost of a decoupler. if your probe is just going to float there and won't move around, it's just a matter of looks: a satellite usually looks better on its own than with a big upper stage attached to it, but it's not necessary and doesn't bring any actual benefit.

[Foxster]

IRL you don't want a big fat fuel tank and engine attached that you have to drag around whilst doing station-keeping manoeuvres. It would be a waste of dV and make the craft unstable. In KSP though, you don't need to do station keeping so it doesn't matter. 

Edited March 13, 2017 by Foxster

[Streetwind]

In real life, there's such a thing as "stationkeeping". That is the act of using RCS to counteract the natural instability of orbits in a real N-body gravitational system. If your satellite doesn't actively maintain its orbit, it will be perturbed by Jupiter, by the Moon, and by local variations in Earth's magnetic field caused by local differences in surface material. (You get a different amount of gravity over the Sahara desert than over the Himalayas, or the North Pole. So your orbit starts to wobble.)

In KSP, we don't have any of these things. But if we did, it would make sense to ditch the final propulsion stage, because that would be extra mass that the RCS system needs to keep forced into the inbtended orbit. You'd shorten the lifespan of the satellite by not decoupling the last stage, because it would have to consume that much more propellant for stationkeeping.

[Red Iron Crown]

A few reasons that upper stages are separated from satellites in real life:

• The extra mass and offset of the CoM is detrimental to the attitude control systems, whether RCS or reaction wheels.
• That mass also reduces the delta-V of the sat's propulsion system, which means it can do fewer station keeping or orbit adjusting maneuvers.
• The stage casts a shadow that can interfere with solar power production.
• The stage adds significant drag. The atmosphere at orbital altitude is awfully thin but not completely ignorable, additional drag means station keeping maneuvers are more frequent or greater in magnitude (or both).

The last one doesn't matter in KSP as the atmospheric pressure is truly zero in space, but the others matter (if less so than in real life).

[kerbiloid]

Unless they call it Agena or Transtage. Then it can stay connected.

[Snark]

4 hours ago, Red Iron Crown said:

The extra mass and offset of the CoM is detrimental to the attitude control systems, whether RCS or reaction wheels.

This one's also a significant difference between real life and KSP.  Real-life spacecraft generally need to spend RCS fuel not just for stationkeeping, but also for periodic attitude control, since real-life reaction wheels are just "storage batteries" for angular momentum-- they can't actually create or destroy it.

As opposed to KSP, where, 1. the minor forces IRL that can cause attitude drift don't exist, 2. ship rotation is killed to precisely zero every time you switch ships or toggle timewarp, and 3. reaction wheels are physics-violating infinite sources of free angular momentum.

So, there's one more reason why this is a thing that IRL ships need to do, but KSP ships don't.

 

[NFunky]

Thank you guys for all the quick answers.  I really appreciate it, and everything you all have said makes complete sense.  Does anyone know if there's a mod that adds station keeping to KSP?

Honestly, I got to wondering about this question while recreating the early Soviet Luna missions.  Those probes were just metal spheres with antennae and basic scientific sensors.  No RCS or attitude control at all I belive?  Certainly not much in the way of station keeping, since they were spending most of their time WAY out in cislunar space.  I'm pretty sure that Luna 1 didn't even separate until it had left Earth's SOI, but it did still separate.  Even Sputnik separated from its booster, but I'm guessing that was to reduce drag and extend orbital lifespan.

I need to do more research myself obviously, but does anyone have any insights into this?

[Streetwind]

16 hours ago, NFunky said:

Does anyone know if there's a mod that adds station keeping to KSP?

There is one that I know of:

The title says 1.1.3 but you can find a link to a 1.2.2 version on the last page of the thread. Can't tell how well it works, I've never used it myself.

[Superfluous J]

One more reason NASA et al dump the final stage and we don't (or shouldn't, whatever) is that NASA plans missions to a very high degree. When they're almost in orbit the insertion stage is actually out of gas, or nearly so. We (generally) just toss stuff together until it works.

[NFunky]

Thank you so much!  I'd seen that mod before, but it wasn't on CKAN for 1.2.2 and the thread does say 1.1.3, so I had no idea it was compatible!  Seems to work fine so far, but I haven't done any rendezvous or anything with it yet.