0x0h

By pulling the extended crew module, drastically overfueling the external tank and igniting the OMS engines at launch, I managed to get my shuttle to lift the heavy-plus target goal. That's pretty impressive considering it was only designed to lift about 20 tons. 3 engine SSME cluster, 2 (liquid fuel) 'OMS' engines, and clustered SRB boosters took me to ~128.5KM circular orbit. Also managed to glide back home even with the compromised Aero/COM from removing the crew module of the orbiter, although I missed the runway by about a KM. Youtube video of the launch/landing. Craft files.

I'll toss my stock and cargo bay lacking into the ring in hopes that it inspires more people to tinker with staged burnouts and SRB clustering. I'll also toss up a quick bonus shot of the mod craft version that does have the bay just because it's a whole lot nicer looking in orbit. -Orbiter (+50) -external tank + boosters (+20) -only solid fueled boosters (+10) -no engines at the external tank (+15) -3 main engines at the orbiter fueled by the external tank (+10) -2 OMS engines at the orbiter to do the Orbit Insertion (+15) -at least 3 kerbal-crew (+10) -orbiter has delta-wings (+20) -orbiter has landing gear (+15) -orbiter lands either at KSC or Island Airport (+5)

Tweakable parts have made this much easier then it used to be since you can now selectively add ballast around the ship without going through the hoops of enabling/disabling the tanks. Plus the benefits of being able to scale back SRBs and limit liquid-fueled engine throttles independently cannot be overstated since if you're going for an STS style of ascent then you've only got a very slim margin for COM drift before things go bottoms up. The difference between one or two percent throttle on an individual engine in a cluster may not sound like much, but depending on placement it may mean several degrees of overall movement. As for shuttles in general, I used one regularly from .18 to the end of .21 for things like space station crew transfers and launching satellites. Mine was a Buran-style based on the Mk 3 with Echo 8's old bay so it's relatively small (1-4 tons of payload + ~20 ton fueled orbiter to LKO) compared to some of the ones I've seen lately, but it was fairly reliable considering how impracticable even real shuttles are. More recently (in no small thanks to tweakables,) I've finally managed to convert it over to a true STS launch system with stock SRBs and engines only on the shuttle itself while remaining stable enough for Mechjeb's auto ascent to fly it to space with almost no added input. Personally I think what makes building shuttles so much fun to build in KSP is that by their nature they're actively trying to murder you the whole way into space, so even when you do find a design that works well you'll almost always have more room for further improvement.

For me, each launch gets named with a unique mission identifier that's independent of lifter and payload that's composed of the launch site, target planet, number of the target satellite (which is omitted if not visiting a moon), and mission purpose. The three letter code then has the launch number appended to the end. In the case of a sub-mission, like a lunar landing by a detached craft, a suffix letter is added to the end of the launch number and is treated as a separate mission for it's duration. Big projects also get a 'common name' with a number designation. For example, UK1X-28 Pioneer 3 means Launch Site 1 (KSC-1), Planet Kerbin, Moon #1, Experimental - Total Launch #28, Pioneer Project Launch #3. Lifter and payloads are also coded with Model/Revision and a common name, and everything is all neatly placed into a log file with mission reports. The whole thing sounds like a nightmare but it's really not that bad. Or at least that's what I keep telling myself that.

Glide landing with a light (or jettisoned) payload is possible, but you have to abuse the flight physics a bit to make it happen by adding more wings for lift and carefully balancing the COM and COL. Having the ability to dump excess stored fuel or ensuring that your de-orbit burn empties the tank is also super helpful, since once you're coming in that fuel is not only dead weight but also affecting the COM of orbiter. Just like the real shuttle, if your balance is off by more then a little bit everything is going to end badly. Anyway, money where mouth is. This is my shuttle, shown carrying a Minimus-bound Remote-Sat to orbit: Not stock, but pretty close. Uses the old Echo8 cargo bay that was patched to the new standard, a modified Mk 3->Mk 2 Adapter that holds liquid fuel and a fuel/RCS-dump toggle, and a Probe-body at the head that's been changed to hold 3 additional crew. She'll also take anywhere between 1.5-4.0 tons to LEO and is capable of bringing about 1 ton back if necessary. Glide only landing once deorbited, and capable of coming down from an inclined orbit. I spent way to much time on this, but it works surprisingly well considering how much of a pain shuttles are in KSP.

Being able to move the thrust vector of the engines in flight and controlling thrust levels on an individual basis would be the best approach, but it is also possible to make COM shift using properly timed firing and shutdown of engines. That said, any shuttle-like design that's not prepackaged in a mod pack is going to have concessions to make. Stock? Forget it if you want a cargo bay or any fuel in space (lookin' at you, Mk 3 Fuselage). More then 3 crew? Have fun fitting a 2M round peg in a ~1.3m square hole. SRB boosters? Not unless you mod the heck out of them. True to the mechanical design of the shuttle? Prepare for it to look like something that came out of a blender. True to the aesthetic design of the shuttle? Three LTV45s aren't going to get you far once your boosters burn out and you've got to lug that external tank. Me? I'm happy with something shuttle-like that's pleasing to the eye and somewhat functional.