Has anyone ever manually flown a spacecraft to orbit, Kerbal Space Program-style?
I have never heard of such a thing, at least on the US side. I'm pretty sure it would be legendary had it ever happened. The Apollo astronauts had that option on the Saturn V. Gene Cernan once said:
You almost wish you hadda guidance failure at liftoff. Because I knew I could've flown that big Saturn V into orbit goddamn near as good as the computer.
I believe manually flying a rocket to Earth orbit from Earth's surface would be impractical for the following reasons:
All current rockets pull massive G forces in the early stages of ascent. It is essential to get up to orbital velocity as quick as possible, because at low speeds you waste propellant fighting gravity. high G forces would make the rocket difficult to control.
Most rockets go into a low earth parking orbit initially, and this barely skims the Earth's atmosphere. Without significant guidance, a pilot would not have the precision to enter such an orbit, so would need to overshoot, putting the spacecraft into a higher (or more likely, elliptical) orbit as this would be a safer option than undershooting and ending up entering the atmosphere and crashing in an undetermined part of the world. Such an orbit would be a waste of propellant. If you are going to provide a guidance system, you may as well link it directly to the engine rather than having an unnecessary step of using the pilot.
Manual orbital rendezvous has been tried, and the first attempt failed due too poor understanding of orbital mechanics https://en.wikipedia.org/wiki/Space_rendezvous
More spectacular manual piloting occurred during the Apollo 13 lunar mission includin changing the orbit. https://en.wikipedia.org/wiki/Apollo_13 . When the service module failed, the astronauts were forced to abandon their descent to the moon's surface and instead use the lunar module as a lifeboat. They executed several burns of the lunar module's engine in order to return to Earth, according to instructions from Mission control and using the earth and celestial bodies out of the window for reference. https://ntrs.nasa.gov/search.jsp?R=20090026451 gives a good overview of what was required. Things to note are:
- Although this was an emergency, they had a lot more time to navigate than they would in a liftoff from earth
- Mission control deliberated quite a bit on each burn before instructing the astronauts on how to execute it. The accident had occured even before the command/service module reached the moon, but they still needed to consider the target landing zone before the first mid-course correction burn. In between burns they tracked the position of the spacecraft to determine the magnitude and direction of the next burn.
- Due to venting of the damaged oxygen tank the astronauts had difficulty identifying stars to use as reference for the burn, but instead used the sun, moon and earth.
- Although there was some manual control, the guidance system was used to help with targeting. A wristwatch was used to check the burn duration
This gives an example of what is involved in navigating a spacecraft. Note that they had several very small correction burns to carry out and given that they had several hours between the burns they had (some) time. I do not believe it would be a good idea to carry out such manoeuvres manually in close proximity to the earth, as an incident would be disastrous.
Shuttle had the capability to fly manual ascents, and the crew was trained to do it. It never happened in the real world.
This excerpt from the Ascent Checklist shows one of the places where the crew could be directed to take over manual control (CSS = Control Stick Steering) and manual throttle control.
The Vostok was controlled completely automatic or from ground, since it was feared that Gagarin might pass out or would be otherwise unable to control the spacecraft correctly. The controls were even locked but Gagarin had a code he could have used to unlock them.
Mercury-Redstone 3 had an automatic ascend but Shepard took manual control during the flight to do some experiments. NASA scientists/engineers didn't even want to include manual controls for Mercury but the pilots insisted they be added (and a window) so they can do something if the automation fails (which turned out to be a good idea).
So even for the very first space flights, in a time when the spacecrafts didn't even have a programmable computer, the launch phase was automated.
Depending on your definition of orbit, the X-15 Aircraft was flown above the Kármán line to 108km in flight 91. At its Apogee, it had no aerodynamic control and was in a rather elliptical orbit. If you insist on once-around, the other answers are most likely the authoritative ones.
In the X-15 flight 51 Armstrong used the MH-96 was an experimental adaptive controller to fly by hand to 207,500 feet (63km) and bounced off the top of the atmosphere when trying to re-enter.
There is both an automatic and manual flight control for the X-15. I didn't find an absolute reference as to whether flight 91 was piloted in manual mode or not. However, some non-atmospheric X-15 flights clearly were in manual mode.