One-Part or Two-Part Spacecraft?

[JonathanDahlq]

Hello everyone.

In real life space agencies often split spacecrafts (for example Apollo) into two parts: a lander and an orbiter.
Do you think that is necessary in KSP for any manned landing with return at all? If so, what planet/moon are you thinking about?
What do you prefer & what do you usually do?

Edited December 4, 2016 by JonathanDahlq

[Gaarst]

Try to direct ascent off Eve and you'll see.

Apollo style crafts are useful for the places that require the most dV:

• Tylo and Eve, because your lander will be so big you can't afford bringing the fuel needed to go back to Kerbin down with you
• Moho, because your orbital craft will be so big you can't afford bringing the fuel needed to go back to Kerbin down with you

Other than that it's not necessary at all, but it often ends up being the lightest solution.

[GoSlash27]

^ What @Gaarst said.

 Not absolutely "necessary" in most cases, but it can be advantageous. Especially in career.

If you have an orbiter that acts as a "mothership/ tanker" and a fully reusable lander, you change your mission from a simple landing to an full-blown expedition. It allows you to collect science at many different biomes instead of just one.
 That's a powerful advantage, considering that interplanetary missions are measured in years.

Best,
-Slashy

[Starman4308]

I doubt it's ever 100% necessary, but it does help quite a bit, particularly for Tylo and Eve as mentioned.

The advantage of having a separate lander is that you get to leave all the return fuel in orbit, whereas a direct-ascent profile requires that you beef up the lander to carry that fuel down to the surface and back up again. The more delta-V you need to get down and back up again, and the more landings you plan to make, the more it favors having a separate lander.

Granted, if you're trying to land on Gilly, the delta-V requirement to land is so trivial that it's a waste, as you're spending a lot of mass on duplicate command modules, engines, docking ports, etc. It more or less comes down to the question of "what's worse: duplicating components for a separate lander, or hauling your return fuel to and from the surface?".

[juanml82]

There is also the issue of designing a ship to land, specially into a body with an atmosphere, versus designing a ship which is never intended to touch the ground.

A lander must be stable on the ground. If it's for a landing within an atmosphere, it needs a degree of heat shielding and it's good to add parachutes to it and it needs an aerodynamic shape. It needs ladders so kerbals can go out and return. It may have wheels to travel around. It can not land on a docking port, or in its engines, for instance. It also needs a certain amount of thrust.

OTOH, a ship which is meant to stay in space can have whatever you want at it ends. You can have exposed gigantors, for instance, because it's not going to burn them on reentry. It doesn't need ladders, it doesn't need an aerodynamic shape (well, you need it to launch for Kerbin, but you might do with a fairing, or you might be assembling in orbit). It doesn't need a minimum thrust.

[Snark]

2 hours ago, juanml82 said:

There is also the issue of designing a ship to land

^ This.

Aside from the design points that juanml82 points out, there's also the matter of TWR.  A ship that just orbits and never lands can have a really low TWR, which means it can run the very high-efficiency, low-thrust engines such as LV-N or ion engines.  Landers need more TWR.  For example, suppose I want to send an unmanned probe to land on Moho, and I want to keep it small and light.  I can use an ion-powered transfer stage that takes it from low Kerbin orbit to low Moho orbit-- ions are great for scads of dV.  But it's pretty hard to land on Moho with ion power; there's just not enough oomph.  So I can have a little lander probe with an Oscar or two and a Spark, which has plenty of TWR.

With KSP 1.2, there's another benefit to having a two-part spaceship, if you're sending your initial pioneering ship to a planet for the first time:  by leaving one part parked up in orbit, it can be a communications relay for the lander.  This comes in really handy-- an orbiter can see more sky than a lander can, and greatly improves the chances of maintaining contact.

If you're planning on doing lots of biome-hopping on a vacuum world, and the biomes are fairly far apart, then an orbiting tanker makes a lot of sense.  The dV you need to launch into a long-distance suborbital hop is not all that different from the dV needed to launch into a low circular orbit.  Call that amount X.  If you start from an orbital mothership, go down, land, and then launch back up again, your lander needs to carry 2X worth of fuel for each hop.  But if your lander doesn't dock-and-refuel each time, then it has to carry enough fuel for all the hops, which is really inefficient-- you're toting tons of fuel through multiple hops, and the rocket equation will bite you on the keister.

 

[GoSlash27]

https://youtu.be/nMe7dRoPRVU?t=125

"Getting to the moon is going to be all about weight. Do you really need to take all of that to the surface?"

 

[wumpus]

Note that for Mun landings, this is probably overkill.  You could always leave a fuel tank + octo probe + solar panels +docking port in Munar orbit for that bit of fuel, but that's about it.

On the other hand, for some reason I built a craft with an LV-N engine for Minums (or the Mun).  It was too much, so I brought it all the way back to Kerbin and left it in orbit.  I then proceeded to use it as a "local shuttle" to Minmus and Mun.  It was an extremely handy two-part ship (where the lander often brought fuel and staged after docking).

[LordKael]

I will occasionally even have 3 part ships. Have a mothership that provides extremely efficient thrust, refining, and power, a reusable lander, and a mining set up. This lets me conduct suites of missions within a single transfer/return window.

[SkyHook]

There are several advantages to having two craft instead of a direct descent craft:

1. Specialization. Jack of all trades, master of none is fine, but specializing allows you to have two craft that are GREAT at one thing vs 1 craft that is OK at two things.

2. Reusability. If you build a great lander, why not leave it in orbit and reuse it for the next time you visit? This saves you money every time you visit the body, and allows you to spend more money up front for a superb design, knowing that it will be reused.

3. Efficiency Why bring all of the mass that you need to get home (heatshield, NERVAS, etc.) to the surface and back? Why include landing legs and a ladder on a craft that is supposed to stay in orbit?

There are more things, but I can't remember! Long story short, build two craft if you can. it will be EASIER and not harder.

Edited December 5, 2016 by SkyHook
Added #3

[Mastikator]

Generally I've found that having an ISRU to be easier than dealing with having a lander.

[Wjolcz]

The size of the solar system and the lack of LS make me not bother with any kind of Apollo style landings. I tried to do these types of missions in the past, but it was nothing but hassle. It's easier to manage a one pod+one tank+one engine lander (which also is a return vessel at the same time).

At least that's what I do on my trips to the Mun. Eve? Never landed anything manned there. I don't think landing there, planting a flag and getting back just for the sake of it is really worth it.

Edited December 5, 2016 by Veeltch

[Waxing_Kibbous]

Bradlet Whitstance uses multiple vessels in some of his videos, it's an interesting technique:

 

[Superfluous J]

Like many here, I use 1 for most things in Kerbin's SOI (Mun and Minmus) so long as I'm doing a simple go-and-return mission. Projects (ISRU, base building) are multiple ships almost out of necessity and/or by design.

Anything outside of Kerbin's SOI gets multiple ships, except maybe a there-and-back to Gilly or something. Ike you could do with 1 ship, but if you're going to Ike you may as well bring enough to do Duna, and IMO Duna warrants a separate ship. Or even better a plane.

[wumpus]

19 hours ago, SkyHook said:

2. Reusability. If you build a great lander, why not leave it in orbit and reuse it for the next time you visit? This saves you money every time you visit the body, and allows you to spend more money up front for a superb design, knowing that it will be reused.

While in general I'm less impressed by this justification, I'd like to point out that the Mun has 17 biomes, and while I'm sure hittting two of them in one landing (typically in and just out of a crater) is fairly easy, getting much more than that will likely require refueling.  This requires a lot more dedication to landing skill than I have, but I've opened up at least half the map by doing similar on Minmus (9 biomes, one refueling trip).

Visiting 17+9 biomes means you can build a rather comfortable rocket to Duna and beyond.  Just understand that all your early launch windows will have passed and you will have to figure out from there.

[Sharpy]

Personally, I'm finding atmospheric reentry hardening to be heavy, not fitting most designs, and bothersome. So I make all my craft that don't go to atmospheric bodies not fit for Kerbin reentry.

Instead, I send a 'dropship' into LKO, with a bunch of reentry capsules, tiny crafts that can do little beyond reentry, but they can do it well. And all returning craft just parks in LKO to form a 'museum', and kerbals use the capsules to land.

[septemberWaves]

I tend towards using separate landing vehicles. I'm not in the habit of performing direct ascent missions, especially outside of Kerbin's SOI. I'll usually make a reusable mothership for interplanetary destinations, and have separate landing vehicles (reusable whenever possible) that are shipped to the planet when needed and stay in that SOI. I enjoy the complexity, and it's quite convenient for extended exploration programs to have a mothership that's fully reusable and can carry a lot of cargo.

[SkyHook]

Definitely depends on your priorities for sure. I am doing a reusable career so this justification is paramount for me, regardless of other considerations. True, it does require more skill to do precision landings and rendezvous/dockings, but reusing craft saves money and makes me happy, so I do it.

[NSEP]

Direct ascent in the stock solar system is waaay easier than a 2-part spacecraft. Exept for Eve, Tylo and Kerbin.

[flywlyx]

It really depends on your docking skill. If docking is not a big problem, you could even dock the ship without RCS, 2 piece is a better choice for most cases. Otherwise, single ship is a better choice.

[Freshmeat]

For a simple Mun mission in early game, a lander is definitely overkill. However, I have started to use Surface Experiment Pack, which requires an engineer and a scientist, as well as roughly a ton of experiments. So my current career Mun missions are full Apollo with a two stage lander and a CSM at a combined mass 20 tons, 150.000 kredits a mission. It might not be the economic way, but so much more fun.

[bobcook]

Sometimes its just fun to build an Apollo-style "Mun orbit rendezvous" mission with a separate lander. Not strictly necessary, as pointed out earlier, but fun. Here is one that I did a couple of years ago: http://imgur.com/a/kAVQq

I find this sort of thing challenges me to improve my rendezvous and docking skills. I can do it pretty efficiently now, and keep getting better every time I do it.

[foamyesque]

It depends on the mission profile exactly how you're going to do your staging, but the golden rule of never taking mass any further than it will be used for applies to everything, and that means an orbital rendezvous is almost always a mass savings (maybe not on, like, Gilly). Even on the Mun or Minmus; your OR target can easily be nothing more than a tank of fuel you then don't need to bring to the surface and back.

[wumpus]

14 hours ago, foamyesque said:

It depends on the mission profile exactly how you're going to do your staging, but the golden rule of never taking mass any further than it will be used for applies to everything, and that means an orbital rendezvous is almost always a mass savings (maybe not on, like, Gilly). Even on the Mun or Minmus; your OR target can easily be nothing more than a tank of fuel you then don't need to bring to the surface and back.

You need to haul at least two clamp-o-tron juniors into orbit, and one all the way down the gravity well and back.  If you want that tank to reliably be there (and not be deleted as debris) you will need an octoprobe as well.  Expect to need RCS on your main ship (and have a use for that monopropellant in the capsule), and a storage bay so your octoprobe doesn't kill the aerodynamics of your ship.

[2] Clamp-o-tron jr: .04t
Octoprobe: .1 (OKTO 2=.04t)
4 RCS thrusters (on the lander so you can dock) :.2t
service bay: .1t
total = .44t

Total amount of delta-v needed to go from Minmus orbit to Kerbin ~100m/s (probably less.  Escape velocity is ~1.4 times orbital velocity, and my handy (.9) map says Minmus has an orbital velocity of 180m/s).

I think the only way you are justifying taking that much extra mass to Minmus for >100m/s delta-v is to stick a couple of solar panels (probably needed but not included, but the weight is pretty trivial anyway) and antenna and call it a communication repeater.  The Mun has an orbital velocity of 580, making it a much better justification (especially if you put on the communication antennas).

[foamyesque]

Nah, the only thing you need is the two docking ports. The probe core's unnecessary, the service bay's unnecessary, and the RCS is unnecessary, at least for super light vehicles b/c you can perform the docking manouvers entirely on your lander's engine and reaction wheels. And on heavier ones, the docking mass overhead is a much smaller percentage, so the fuel savings become more significant.