QF9E

Sorry to gatecrash this challenge with a KSP1 entry (I haven't bought KSP2 yet), but I thought this too funny not to share. Would this method work in KSP2 as well?

Yes, I tried it myself before posting. You can use ore as ballast, as ore is denser than water. You need quite a bit of it, as rover wheels (at least the large rubbery black ones that I tried) provide quite a bit of buoyancy. Mine was slow as hell, though, and it had serious problems going uphill because it was so heavy.

Build a rover that is able to drive on the sea floor?

Grinding the biggest Crack of them all... Unfortunately it appears to be very hard to approach the bottom of the Dres canyon in a stable orbit, so the minimum altitude of about 60 meters above ground level is reached quite a bit before entering the canyon.

It turns out no steering of any kind is necessary. I launched straight up at dawn during a Duna transfer window and when Kerbin passed through the descending node with respect to Duna to land a capsule on the Duna surface without the need for a separate trans Duna Injection burn or any mid-course corrections. Launching straight up at dawn means you're launching into Kerbin's prograde direction around the Sun, which is what is needed for a Hohmann transfer orbit to Duna Launching when Kerbin passes the descending (or ascending node) means you won't miss Duna half an orbit later by flying over or under it. It did take some fine-tuning of the exact launch time to get this to work, but after half a dozen attempts I found that for my rocket, Year 5, Day 366, 4 hours 08 (which is just before dawn at the KSC) works fine. Full mission report here: https://imgur.com/a/Y9QpQsg

Decouplers exert a force on your craft during decoupling. So you could conceivably mount some decouplers around your craft and steer with them.

What, exactly, is your definition of ISRU? Would it be allowed to fill some ore tanks in the VAB and convert that ore into propellant on the fly, as long as you don't mine any ore?

I think one could pretty much fly endlessly in the upper atmosphere, by making a craft that has a fairing as its root component, and fully occluding the forward and backward nodes - that way it will not generate any drag when flying prograde. As to how to get it into a 49.9 x 49.9 km orbit: from an orbit just above the atmosphere (I started with a 50.001 km circular orbit) lower PE into the atmosphere. While in the atmosphere, fly with the nose pointing in the normal direction (i.e., sideways on, so that the craft does induce drag) until drag lowers AP to 49.9 km and then turn prograde to cancel all drag. I tried it (although I cheated my craft into Duna orbit) and completed a full orbit in this way without AP or PE degrading noticeably.

Another suggestion: include the pods as external payload attached to the external tank. Designs for this existed for the actual Space Shuttle, see, e.g., https://www.aiaa.org/docs/default-source/uploadedfiles/about-aiaa/history-and-heritage/shuttlevariationsfinalaiaa.pdf?sfvrsn=b8875e90_0

Yup. I've been working on an Eve SSTO for a couple of years and finally succeeded earlier this year: https://imgur.com/a/1AD3u40 . I did need a support package though. My Eve SSTO is barely able to make Eve orbit and has to be topped up from a tanker before returning to Kerbin.

@OJT Awesome Eve STS-2 mission. Landing on Eve and getting back into orbit isn't easy at the best of times, but doing it with a lander that fits inside a Shuttle bay is really quite hard. Congrats on pulling it off successfully, and from Eve sea level no less! Next challenge: Eve STS-3: land a Shuttle on Eve and return it safely to Kerbin ;). For the commander level: land at less than 1000 meters above sea level on Eve, and land the Shuttle on a runway.

Ok, I did look it up. According to the KSP wiki it is not possible to land on Jool: Source: https://wiki.kerbalspaceprogram.com/wiki/Jool. This matches my own experience in trying to land on Jool. My own experiments went as described: the craft will explode once you reach -250 meters, although I haven't experienced glitches or corrupt saves as a result. As to the note on "prior to version 0.23": Version 0.23 was released back in 2013. I don't think many people will be willing to play such an old version of the game, just to be able to complete this challenge. That said, by all means prove us wrong. Fly your own mission where you land on Jool, take a surface sample, plant a flag and get back to Jool orbit, and present video evidence of your mission here.

At 1:47 in your video, at about 10k or 11k altitude you give a series of yaw inputs to the right, as is indicated by the gizmo in the lower left corner. This dips the nose of your craft below the vertical, initiating flight with an angle of attack on the wing. Which in turn initiates an uncontrolled roll, which ends when your craft is flying with its wing vertically and towards an azimuth of about 160 degrees, at about 1:51 in the video. Please don't presume to know what I did or did not do. You have exactly zero evidence so your opinion on the matter is entirely baseless. That does not make any sense. Of course I try to keep my vehicle under control. You can use the airbrake part to lift your rover off the ground as @miklkit describes in their post. You can also use an airbrake to right a rover that has turned over. We are well aware what you are talking about. What you are apparently unaware of, is that it is possible to land a Kerbal on their own from low Kerbin orbit. Or build a small craft based around a command chair and re-enter the atmosphere with that. In cases such as these, a personal chute comes in handy to land your Kerbal safely. As to "but even in a returning craft from space there is no use as it either lands or too fast to bail": Yuri Gagarin, Gherman Titov, Andriyan Nikolayev, Pavel Popovich, Valery Bykovsky and Valentina Tereshkova would like to have a word with you. In case you don't recognize at least some of those names: those are the Vostok cosmonauts, Vostok being the first Soviet crewed spaceflight program, somewhat akin to the US Mercury. They ejected from their capsules after re-entry and landed under their own personal parachutes. This can be done in KSP as well.

That's not what is going on: 1. The gimbals of your engines are not balanced. This induces roll when you give pitch input when you initiate the gravity turn, which makes your craft hard to control. 2. Flying with a non-zero angle of attack on the wing panels inside the atmosphere induces lift. And since the center of lift is a long way away from the center of mass, this lift induces so much roll torque that SAS is unable to compensate. Your craft ends up with the wing oriented vertically because of this. You also end up no longer flying to the East, and trying to compensate by giving yaw input again results in a non-zero angle of attack and uncontrollable roll. 3. During your circularization burn the craft was clearly out of balance. Note that the pitch indicator at the bottom left of your screen is way off-center, which means SAS must work overtime to keep the craft on course. Also note that once the pitch indicator is fully at the top or bottom of its range it cannot control pitch anymore, as your engine torque is by then more than the reaction wheels in your craft can handle. If you fly the same rocket model 90 degrees rotated on the pad to what you have been doing, you can avoid 1 and 2, and you can even control your gravity turn using differential thrust. I got a close copy of your craft into a 75 x 75 km orbit with over 600 m/s dv to spare without too much effort and without ever losing control.

Please indicate which of my suggestions amount to brute force. It is entirely possible to build and fly a Space Shuttle without using excessive RCS, if that is what you mean. That said, you need the RCS system anyway if you want to be able to maneuver your shuttle accurately in space. If I may offer you some personal advice: please refrain from posting stuff like this. It's no use boasting about your achievements, and it may well result in unnecessary friction with other forum members. Real life space missions also require RCS and mid course correction burns (MCC). Of course, these are usually planned meticulously, but occasionally there has been some eyeballing in real life. Apollo 13 comes to mind, where they performed an MCC entirely by hand, because the flight computers were offline to save on electricity. Another, less successful example is Gemini 4: it attempted a rendez-vous with its own upper stage after launch, but failed to do so as the astronauts tried to eyeball the maneuvers and wasted most of their fuel. A more recent example is the Nauka docking with the ISS, which was "eventful", for lack of a better word. That said, if you use the maneuver planner to plan your MCCs in KSP, you don't need any eyeballing beyond switching the engines on and off at the appropriate times. Moreover, if you own the Breaking Ground DLC you can use the KAL-1000 to perform a very precise and entirely computer-controlled launch: While this is not a Space Shuttle, I don't see why you couldn't use a KAL-1000 to balance the thrust of an asymmetric craft. However, it does take a bit of patience to program the KAL-1000 just right. Which, given the amount of man hours spent on real-life spaceflight avionics, is quite realistic. If you want even more computing power to control your craft, I can recommend the kOS mod. That mod offers a fully programmable flight computer inside KSP.