cantab

Looks very well presented. A couple of things I've noted: You've got points for synchronous orbits around bodies where it's not possible (because the SOIs are smaller than the synchronous orbit radius). If you want to leave them in and have the students discover the impossibility themselves, that's fine, otherwise I'd blank them out. Since you're using cost, if you haven't already I'd rework all the part costs. At present they're all over the place and make things weird. You might want to look into how things are for real rockets.

Yeah, to plant a flag needs a Kerbal, making this challenge boil down to: Land a Kerbal on every planet (even Jool, which is possible, but glitchy as heck), and as many moons as possible, and make a probe sundive - all from one launch. It's perfectly possible - anything's possible if you've got a big enough rocket - but even without returns it's a big mission.

If it's with the new parts, then it's down either to having a greater TWR at liftoff, or possibly to differences between vacuum and atmospheric Isp and thus delta-V. The new four-engine cluster has a relatively small difference between its performance in vacuum and its sea-level performance compared to most older engines, which will result in using less vacuum delta-V, which is probably what you're reckoning things off. By contrast the KR-2L has a very large difference in performance depending on atmospheric pressure, so if you use it as a first-stage engine (which it's not meant for, but it looks the part for many rockets) you'll use more vacuum dV.

You are fixing the staging before launch, right? Since your screenshots show the engine ignition and chute deployment in the same stage. Also, you'll really want to make sure the thrust axis points through the centre of mass. Easiest way is to duplicate that part on the left and put one on the right too. One other thing is the cupola can be troublesome, it's heavy and extra-draggy. Try swapping it for a different command pod.

Later tonight I found that part clipping makes rockets finicky. The VAB kept eating the fuel lines to my second stage's outer engines, and my first stage kept losing an engine without explanation - nothing shown as breaking, just the engine not firing and reporting itself oxidizer-deprived. I threw the first stage out and checked the second stage works, at least. Will have to fix the first stage tomorrow. Failing that, anyone know a mod with 7.5 metre radius tanks?

Engineer Redux and Precise Node. Candidates for the near-future: MechJeb. My first use case for it being to reproducibly test launchers. My own ascents are good enough for flying around, I always build in a dV margin anyway, but not consistent enough for properly comparing designs. An aircraft "set", with FAR (of course), Procedural Fairings or similar, and probably Firespitter. Perhaps something with bigger tanks, KW or Novapunch I believe. I was building a Saturn V analogue earlier today and started wishing it was actually round.

No. I am, however, doing Sandbox at the moment. (Well, career hacked to be like Sandbox but with SCIENCE, but same difference). I'm certainly holding off playing career mode until the .24 update at least.

I got a reminder that the large ASAS unit isn't very strong (It's near the join between the grey tanks that are the rocket, and the half-full orange ones that are the test payload.) With predictable results

In terms of studying the Moon itself, loads. However, that can be done fairly well by unmanned landers and rovers, especially as a lunar rover can be driven in near real-time and so travel much faster than one on Mars or Venus, though it may still be slower than a manned rover. Compare Apollo 17's moon buggy, the unmanned Lunokhod 2, and the Opportunity rover: They all drove about the same distance, but took 4 1/2 hours, 5 months, and 10 years to do it respectively. In terms of science to do from the Moon, again plenty. However a practical approach again might be to design the experiments to be autonomous, while having the possibility of manned servicing missions if needed. This worked very well for Hubble, though it is much easier to get to being in LEO. In terms of science to do on people, I think the big one is understanding how the human body handles reduced gravity. We know a good amount about zero-g, and of course loads about 1-g, but nothing of what lies between. Considering the comparison between the Moon and Mars, I think it's fair to say that if humans can survive, stay healthy, get water, and so on on the Moon, we can almost certainly do it on Mars. Meanwhile the Moon offers the advantage of being closer to get to, and possible to get back from in a hurry if needed.

For launching to LKO, the two main approaches are to put a probe core on the upper stage and leave it with a bit of fuel to deorbit itself, or to stop the circularisation burn with the periapsis below 69 km and preferably below 23 km, separate from the upper stage, and finish the circularisation with the payload's engine. (If you have any ship or debris orbiting Kerbin with a periapsis between 23 and 69 km you'll have to "fly" it in order for drag to affect it and bring it down. Same idea, but with different numbers, for anywhere with atmosphere.) For trans-Munar injection, you can make the previous stage hit the Mun, fly past into a wide Kerbin or even solar orbit where it won't hit anything, or fly past on a free-return trajectory and hit Kerbin. Same options for Minmus. For interplanetary transfers you can try for an impact but you'll need to either be precise or have control over the stage, otherwise just leave it in solar orbit. For discarding spent stuff from orbits around bodies other than Kerbin, your best bet is to have a probe core and a little fuel in them for a deorbiting. Next best is to use your active stuff dock/claw and do the deorbit, but I'd only do that if your spent part lacks an engine, otherwise it's more efficient to dock/claw and give it a little fuel. You can deorbit something with a manned pod too, should it lack a probe core. Have a Kerbal do the deorbiting, bail out, and reorbit himself with his jetpack then rendezvous with his craft. EVA rendezvous takes a bit of practice but isn't too hard, you can bring up the map view with M to get near then use the target marker to judge lateral and radial speed and distance to get close.

Yup, welcome to orbital mechanics! Don't worry, with a bit of practice it'll start feeling natural enough. Most things have been covered very well by the other posters. I'll just add: You'll do almost all your rocket flying on instruments - the navball and the map view, basically. You may want to make the UI larger, there's a slider in the Main Menu settings. Keep these two Wiki pages handy: http://wiki.kerbalspaceprogram.com/wiki/Basic_maneuvers and http://wiki.kerbalspaceprogram.com/wiki/Key_bindings

I always have an Abort system set up, but never yet the new LES. My designs so far have always wanted the nose for something else: the asteroid lander on Tycho, the main claw on Macbeth, and a shielded docking port on Chamberlain.

I've never had much luck with aircraft, and am a bit unsure about seriously trying in the stock aerodynamics. (Meanwhile I don't want to find FAR breaks my rockets). That said, I did have some reasonable success with my Chamberlain "space bus". Launches on top of a rocket, crew compartments (but not the engine and fuel tank) come down with wings to allow a controlled glide before a final chute landing. Glides quite nicely in the lower atmosphere, though I still missed KSC by miles on my first go.

You might build an asymmetric rocket if you attached a reaction wheel radially and balanced it with other radially-attached parts. And you're probably more likely than not to build an asymmetric space station. And I can confirm that in .23.5 at least ships are indeed hard to break. That's 4 asteroid control sections each with 4 reaction wheels, connected to the rest of the ship by Regular Docking Port-Quad Adapter-Sr Docking Port both above and below. It bent back and forth on attempting to turn as you can see, but didn't break until I deliberately made inputs to make the bending worse.

I'd think about putting some proper fuel and an LV-1 or 48-7S engine on your rescue pod. It's more efficient than RCS. You're also better off with 8 RCS ports, in two groups of four at the front and back. Just two like you've shown gives you no left-right translation so you'd have to roll the craft. Also, you can save weight by ditching most of the parachutes. One on the nose or a couple of radials should suffice.

A screenshot showing how your orbit will be after it's left Kerbin's SOI would help. If it's still reasonably similar to Kerbin's, it shouldn't be too hard to rendezvous in Kerbol orbit. If you just want to bring the crew back, your best bet is to send a rescue ship with sufficient empty space. Just make sure the seats are empty on launch, if you use command modules they'll fill themselves up by default. Don't bother docking, just rendezvous and EVA the crew over. (As mentioned, you still need to EVA even if you do dock). If you want to bring the station back it's a little trickier. Simplest would be to claw it, refuel its own tank (Alt-Click the tanks in turn and you can transfer fuel between them), declaw it, and have it burn back into Kerbin orbit. However if its own fuel tank isn't big enough to do the task in one go repeated refuels may be a nuisance. An alternative would be to dock a pair of drop tanks to the station's docking ports.

Finally did the Mun flyby in my cheap science mission. Now to work out the course correction for the Eve encounter, and sort the gravity assist off that. It looks like I definitely won't make Moho with a single Eve assist, so I get the extra challenge of lining things up for a second pass without waiting ages. Then I can get on with some asteroid deflection. I've got one due to impact in about 17 days, I'd prefer if it didn't.

Well, I decided to go for getting the maximum science out. The Smith 2 has now completed its Mun flyby. I experimented a bit, and determined that to get the most science I needed to stay in normal time just before and through the critical low part of the flyby, spamming the gravity experiment in order to deliberately backlog it and make sure I didn't miss any small biomes. Temperature doesn't depend on biomes, so I kept that result for later transmission so it wouldn't keep showing when I hit the experiments action group. Then once the science gathering was over I hit the timewarp. Because most of the experiments were still in the queue for transmission, they went out for the full value. The drawback was having to wait a few real-time hours for the transmissions to complete, since they weren't sped up by timewarping. Summary: Cost: $12.480M Experiments: Temperature and Gravity Science: 284.64 Science/M$: 22.81 So far, not the greatest of scores. Actually, the worst of scores. But then this is just phase one of the mission. Even once I make my (rather expensive) mid-course correction for Eve the Smith 2 will still have around 1300 m/s of dV in the tanks. Nonetheless, feel free to stick it down as a Mun mission for now Jasonden. Full photo set on Flickr including some more descriptions than here. Mission cost reminder. Some launch shots. Like the mission emblem on the SRB? Eve transfer orbit including Mun flyby established. Science before. I have been doing other stuff in the save (it's ended up being my main save for .23.5), but nothing from the Mun. Just one of the many science experiments run. It really felt like the probe was going to hit the Mun when it was coming in! Goodbye Mun. Science after.

Not really. It just wants a fairing wrapping round the whole thing and it'd look dead on. OK maybe having 7 third-stage engines is a bit iffy, but that's not the biggest deal.

I'd like to see the biome mechanics for career mode tweaked somewhat first. At the moment it feels like they encourage spamming the same mission over and over. Some sort of diminishing returns would seem appropriate, so players get rewarded more for visiting new places.

Is it, though? Does anyone know what the widest possible orbit is without sending everything haywire?Even if it is finite, remember the system's 3D. Two orbits can apparently cross entirely safely if one is above the other. So I believe planetary collisions are only possible if two different modders happen to copy the same orbital inclination and longitude of ascending node (most obviously if it's a zero-inclination orbit). SOI "collisions" are probably a fair bit more likely, being as planetary SOIs are much bigger than the planets themselves, and I can see them causing weird issues.

Well Wikipedia mentions the ITN being based around the Lagrange points, and the orbits being described by the restricted three-body problem, neither of which we have in KSP which uses the simpler patched conic approximation.

Really looking the part. Can't wait to see it down on Moho.

You realise it's the ARM engines that are closer to realism, right?

Brilliant study, and a great reminder that there are always new interesting things to do in KSP.